Citation
Partners for early intervention and healthy child development

Material Information

Title:
Partners for early intervention and healthy child development linking attributes and characteristics of stakeholder networks to performance
Creator:
Mobbs, Robyn I
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file : ; (226 pages)

Subjects

Subjects / Keywords:
Health services accessibility ( lcsh )
Health services administration ( lcsh )
Health services accessibility ( fast )
Health services administration ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
The reliance upon interorganizational relationships in health and public health systems is extensive; however, there is a shortage of evidence that focuses on stakeholder network performance. Responding to this gap and the call for a systems approach, the study's first research question asks whether community based stakeholder networks operating within the same context exhibit commonalities in network-level attributes and characteristics relative to their performance. The second question focuses on organizational level attributes and characteristics, asking: do the relative connectivity and structural positions of member organizations affect stakeholder network performance? In a mixed method, cross case research design, the characteristics and attributes of five stakeholder networks were analyzed and compared. Whole-network, dyad and organizational measures were examined through network analysis, and analyzed with performance measures based on external stakeholder assessments, network self assessments and local early intervention assessments. Emergent trends were then analyzed through content analysis. Focusing on the dynamic interactions of public health agencies with public and private organizations that affect health, the five stakeholder groups are goal-directed networks working to ensure standardized developmental screening, referral and follow through for children in their Colorado communities. The stakeholder networks were mobilized to overcome the segmented system of early intervention services. The study found that there were no relationships in whole-network level attributes and characteristics relative to performance. Investigation into the second research question revealed a positive relationship between network structures with a system-building organization in a key position relative to network performance; and a negative relationship between network structures with a service provider organization in a key position relative to network performance. There was also a positive relationship between the levels of integration of primary care providers relative to network performance. Results also showed that trust among organizations within a stakeholder network is not a sufficient condition for network performance. The study findings provide new evidence about how different characteristics and attributes of stakeholder networks affect performance, adding to a much needed knowledge base to guide the use, formation and management of organizational networks for strong health and public health systems.
Thesis:
Thesis (Ph. D.)--University of Colorado Denver. Public affairs and administration
Bibliography:
Includes bibliographical references.
General Note:
School of Public Affairs
Statement of Responsibility:
by Robyn I. Mobbs.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
|Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
868159547 ( OCLC )
ocn868159547

Downloads

This item has the following downloads:


Full Text
PARTNERS FOR EARLY INTERVENTION AND
HEALTHY CHILD DEVELOPMENT:
LINKING ATTRIBUTES AND CHARCTERISTICS OF
STAKEHOLDER NETWORKS TO PERFORMANCE
by
Robyn I. Mobbs
B.A., University of Colorado Boulder, 1999
M.B.A, University of Southern Queensland, 2004
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Affairs and Administration
2013


This thesis for the Doctor of Philosophy degree by
Robyn I. Mobbs
has been approved for the
Public Affairs and Administration Program
by
Christine Martell, Chair
Danielle M. Varda
Jessica Sowa
Laura Pickier


Mobbs, Robyn, I. (Ph.D., Public Affairs)
Partners for Early Intervention and Healthy Child Development: Linking Attributes and
Characteristics of Stakeholder Networks to Performance
Thesis directed by Associate Professor Christine Martell.
ABSTRACT
The reliance upon interorganizational relationships in health and public health
systems is extensive; however, there is a shortage of evidence that focuses on stakeholder
network performance. Responding to this gap and the call for a systems approach, the
studys first research question asks whether community based stakeholder networks
operating within the same context exhibit commonalities in network-level attributes and
characteristics relative to their performance. The second question focuses on
organizational level variables, asking: do the relative connectivity and structural positions
of member organizations affect stakeholder network performance?
In a mixed method, cross case research design, the characteristics and attributes of
five stakeholder networks were analyzed and compared. Whole-network, dyad and
organizational measures were examined through network analysis, and analyzed with
performance measures based on external stakeholder assessments, network self
assessments and local early intervention assessments. Emergent trends were then
analyzed through content analysis. Focusing on the dynamic interactions of public health
agencies with public and private organizations that affect health, the five stakeholder
groups are goal-directed networks working to ensure standardized developmental
screening, referral and follow through for children in their Colorado communities. The
stakeholder networks were mobilized to overcome the segmented system of early
m
intervention services.


The study found that there were no relationships in whole-network level attributes
and characteristics relative to performance. Investigation into the second research
question revealed a positive relationship between network structures with a system-
building organization in a key position relative to network performance; and a negative
relationship between network structures with a service provider organization in a key
position relative to network performance. There was also a positive relationship between
the levels of integration of primary care providers relative to network performance.
Results also showed that trust among organizations within a stakeholder network is not a
sufficient condition for network performance.
The study findings provide new evidence about how different characteristics and
attributes of stakeholder networks affect performance, adding to a much needed
knowledge base to guide the use, formation and management of organizational networks
for strong health and public health systems.
The form and content of this abstract are approved. I recommend its publication.
Approved: Christine Martell
IV


DEDICATION
I dedicate this work to my husband Simon. I cherish and adore you and cant
imagine this world without you. I am grateful for your love and enduring support, and I
thank you for coming along on this journey with me... even though it might not have been
your first choice!
And to Emily and Jackson: you are both the center of my universe. I love you
entirely and completely with all my heart. I am grateful for all your love and smiles that
have unknowingly seen me through late nights of studying and researching, and kept me
moving forward on this journey. (Em, I started this PhD the month you were born, and
Jack, you entered this world right in the middle of it). You constantly amaze and inspire
me, and I hope you know that you are capable of anything (with some hard work and
perseverance, of course!). I look forward to every tomorrow with you.
And to my parents, Chris and Kit: you are extraordinary. I thank you from the
bottom of my heart for all your love, support, wisdom and guidance. I am lucky indeed!
And to Ashley: thank you for being the wonderful sister that you are, and for being a
great aunt to Emily and Jackson so that I could occasionally write in the daylight! You
are amazing, and inspire me every day with the knowledge that there is always a way, it
just might mean doing things a little differently! And to Michael: thank you for being an
amazing big brother. You have pushed and inspired me; encouraged and challenged me-
thank you! And to Carolyn and Brian, Keith and Lyn, Catherine and Chris and my truly
incredible friends: thank you for being the wonderful you that you all are! I love and
treasure you all.
v


ACKNOWLEDGMENTS
I would like to thank my amazing committee of incredibly smart, furiously hard
working, and truly amazing women: Christine Martell, Danielle Varda, Jessica Sowa and
Laura Pickier. Christine: thank you for your expertise, guidance and support throughout
this entire process. You have been a solid rock for me to lean on, and a brilliant advisor-
proving invaluable feedback and encouragement when I need it most- thank you!
Danielle: thank you for your phenomenal mentorship in public health systems and
network research. You are a true mentor is every aspect of the word and I am forever
grateful. Jessica: your knowledge of the literature and seemingly photographic memory
of pertinent studies has been an incredible help to me throughout this endeavor, as has
your assistance and support for teaching. Thank you for everything. And Laura: you are
an outstanding and caring physician and colleague, and your dedication and abilities to
improve the lives of those all around you is truly remarkable. Thank you all so much.
I would also like to thank Eileen Bennet and the wonderful and dedicated
members of the ABCD team in Colorado. I am grateful for your time and support. Your
efforts are so valuable, not only to me during this project, but to all the families and
children in the communities that ABCD impacts. Thank you.
I am also grateful for the funding and ongoing support from the National
Coordinating Center for Public Health Services and Systems Research and the Robert
Wood Johnson Foundation. The receipt of the Assuring the Future of Public Health
Systems & Services Research: Dissertation Grant Award was an honor, and provided
invaluable financial support. Thank you to the wonderful leadership and staff at both
organizations for your confidence in my research, ongoing technical assistance, and


wonderful encouragement and inspiration along the way. I feel humbled to be in the
company of such great leaders and scholars in PHSSR, and truly appreciate your time and
efforts. I am also grateful for the scholarship from Academy Health for the ARM. I have
such great appreciation and respect for all that you do in the field and thank you for your
support.
Vll


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION...............................................................1
Background..............................................................1
Health Care and Public Health Systems................................2
Statement of Research Problem...........................................4
Interorganizational Relationships....................................5
Theoretical Framework................................................6
Research Questions......................................................7
Study Design.........................................................7
Plan for Dissertation...............................................10
Key Definitions and Terms...........................................12
Significance...........................................................15
II. RESEARCH CONTEXT: SYSTEM OF EARLY INTERVETION.............................17
Background.............................................................17
Early Intervention Services.........................................17
Early Intervention and the Public Health System.....................21
National Policy Context:........................................21
State Policy Implementation.....................................22
Implementation in Colorado......................................24
Early Intervention and the Education and Human Services Systems.....25
National Policy Context.........................................25
State Policy Implementation.....................................25
Implementation in Colorado......................................26
viii


Limitation in Early Intervention System............................27
Health Care System....................................................29
National Context...................................................29
Implementation in Colorado.........................................30
Segmented Early Intervention System.......................................30
Identification and Screening: Health care system involvement..........32
Referral to Early Intervention........................................33
Provision of Early Intervention Services..............................34
Implementation Challenges.............................................35
Assuring Better Child health Development (ABCD) in Colorado........36
Conclusion............................................................38
III. LITERATURE REVIEW...........................................................39
Introduction..............................................................39
Guided by Theory......................................................39
Research Questions....................................................41
Systems Theory............................................................41
Overarching Theoretical Framework.....................................41
Application to Early Intervention System in Colorado..................43
System of Care Literature.............................................47
Network Theories..........................................................49
Definition of Network.................................................51
Assumptions of Network Theories.......................................52
Theoretical Foundations of the Network Approach.......................54
Organizational Network Research...........................................55
Typology of Organizational Network Research...........................57
IX


Impact of Whole-Network Attributes and Characteristics on Network
Outcomes.............................................................59
Network Structure................................................60
Network Trust....................................................61
Impact of Organizational Attributes and Characteristics on Network Outcomes
.....................................................................63
Organization Positions within Networks...........................63
Organization Roles within Networks...............................64
Connectors within Networks.......................................66
Strength of Ties within Networks.................................66
Network Analysis in Health and Public Health Systems Research........68
Network Performance..................................................71
Conclusion...........................................................74
IV. METHODOLOGY...............................................................75
Research Design.........................................................75
Overview.............................................................75
Identifying the Data Needed.............................................77
Network Performance Data.............................................77
Whole-Network Data...................................................79
Organizational and Dyadic Data.......................................80
Participants and Sites..................................................82
Selection of Stakeholder Networks....................................82
Stakeholder Network Members..........................................85
Census of Ties...................................................88
Human Subjects Review............................................89
Recruitment of Subject Population................................89
x


Data Collection......................................................90
Variables and Measures..................................................92
Network Performance- Dependent Variable..............................92
External Assessment..............................................93
Network Self Assessment..........................................94
Early Intervention Colorado Assessment...........................94
Network Demographics.................................................95
Whole-Network Independent Variables..................................95
Density..........................................................95
Degree Centralization............................................96
Network Trust....................................................97
Organizational Independent Variables.................................98
Degree Centrality................................................98
Betweenness Centrality...........................................99
Brokerage.......................................................100
Relative Connectivity...........................................100
Relationship Strength...........................................102
Data Analysis..........................................................107
Network Analysis....................................................107
Content Analysis....................................................Ill
Study Validity and Limitations.........................................112
Validity............................................................112
Limitations.........................................................114
V. FINDINGS..................................................................116
Performance............................................................117
xi


External Assessment
117
Members Assessment.................................................118
Local Early Intervention Assessment.................................119
Demographic Analysis...................................................124
Demographic Variables...............................................124
Cross Case Analysis: Demographic Variables..........................125
Network Size....................................................126
Network Age.....................................................127
Relative County Size (Population)...............................128
Whole Network Cross Case Analysis......................................129
Whole-Network Variables.............................................129
Density.........................................................130
Degree Centralization...........................................131
Overall Network Trust...........................................132
Network Graphs......................................................133
Cross Case Analysis- Organization Level................................138
Role and Influence of Public Health.................................138
Degree Centrality...............................................140
Relative Connectivity...........................................140
Closeness and Betweenness Centrality............................142
Brokerage.......................................................144
Integration of Primary Care Providers...............................146
Content Analysis.......................................................152
Organizational Perspectives: Systems Building vs. Service Provision.152
Conclusion.............................................................155
xii


VI. DISCUSSION
157
Linking Network Attributes and Characteristics to Performance............157
Summary of Investigation.............................................157
Exploring Research Question One..........................................158
Network Structure....................................................159
Network Trust........................................................161
Exploring Research Question Two..........................................162
Organization Positions and Roles.....................................163
Contributions to the Literature...................................167
Strength of Ties.....................................................171
Conclusions for Early Intervention, Health and Public Health Systems.....174
Conclusion 1: Perspectives of Key Actors Matters.....................175
Conclusion 2: Integration of Primary Care can Strengthen Outcomes...175
Conclusion 3: Bigger Is Not Necessarily Better.......................177
Future Research Needs....................................................178
REFERENCES......................................................................182
APPENDIX........................................................................198
1. Early Intervention Process in Colorado.......................................198
2. Survey Invitation............................................................199
3. Survey Questions.............................................................201
4. Stakeholder Network Self Assessment Responses................................207
xiii


LIST OF TABLES
Table
I. 1 U. S. Health Rankings........................................................1
II. 1 Number of Individuals Served and Annual Appropriations for Title V of Social
Security Act......................................................................24
II. 2 Number of Children Served and Annual Appropriations for Part C of IDEA......26
III. 1 Typology of Organizational Network Research................................58
IV. 1 Summary of Research Design..................................................81
IV.2 Assessed Stakeholder Activities..............................................84
IV. 3 Conceptual and Operational Definitions -Variables and Measures.............104
V. 1 Network Performance.......................................................119
V.2 Summary of Network Demographics..............................................125
V.3 Summary of Whole Network Characteristics......................................129
V.4. Measures of Structural Position and Influence of Network Members.............139
V.5. ECC and PH Relative Connectivity.............................................145
V.6. El Programs Relative Connectivity and Brokerage............................146
V.7 Classifications of Member Organizations......................................153
xiv


LIST OF FIGURES
Figure
II. 1 Systems That Make Up the Early Intervention System at the National, State and
Local Levels....................................................................21
II. 2 Steps in Early Intervention Process, Cross Listed with System............31
III. 1 El System Inputs, Outputs and Outcomes (Colorado).......................46
IV. 1 Screenshot of PARTNERTool Survey Login..................................91
IV. 2 Sample Matrix of Stakeholder Network.....................................110
V. 1 Comparison of Stakeholder Network Performance...........................122
V.2 Stakeholder Networks along a Performance Continuum.........................124
V.3 Variation in Network Performance by Size..................................126
V.4 Variation in Network Performance by Age...................................127
V.5 Variation in Network Performance by Relative County Size...................128
V.6 Variation in Network Performance by Density................................130
V.7 Variation in Network Performance by Degree of Centralization...............131
V.8 Graph of Stakeholder Network Alpha.........................................134
V.9 Stakeholder Network Graphs.................................................136
V. 10 Network Performance by Relative Connectivity of ECCs.....................141
V. 11 Network Performance by Relative Connectivity of Els......................142
V. 12 Network Performance by Brokerage of Els..................................144
V. 13 Graph of Stakeholder Network Alpha, Coordinated Activities...............148
V. 14 Stakeholder Networks, Coordinated Activities.............................150
A. 1 Self Assessment for Stakeholder Network Alpha.............................207
A.2 Self Assessment for Stakeholder Network Bravo.............................207
xv


A.3 Self Assessment for Stakeholder Network Charlie.............................207
A.4 Self Assessment for Stakeholder Network Delta..............................208
A. 5 Self Assessment for Stakeholder Network Echo...............................208
xvi


LIST OF EQUATIONS
Equation
IV. 1 Density.................................................................96
IV.2 Network Centralization...................................................97
IV. 3 Degree Centrality.......................................................99
IV.4 Betweenness Centrality...................................................99
XVII


LIST OF ABBREVIATIONS
ABCD
BOCES
CCB
CDC
CDHS
CDPHE
COMIRB
El
IDEA
IOM
NACHO
NIH
NPHPSP
Assuring Better Child health Development program
Boards of Cooperative Education
Community Centered Board
Centers for Disease Control and Prevention
Colorado Department of Human Services
Colorado Department of Public Health and Environment
Colorado Multiple Institutional Review Board
Early Intervention
Individuals with Disabilities Act
Institute of Medicine
National Association of County and City Health Officials
National Institutes of Health
National Public Health Performance Standards Program
Members of the Stakeholder Networks in this Study:
PH Local
El Local
ECC Local
PC Local
ECC Local
EC Local
BH Local
public health department or agency
Community Centered Board, the El service provider
Early Childhood Council
pediatrician or primary care providers
Early Childhood Council
early childhood centers/preschools
behavioral health center
xvm


CHAPTER I
INTRODUCTION
Background
America aims to be "a society in which all people live long, healthy lives" (United
States Department of Health and Human Services [HHS], 2012, Message from the
Secretary, para. 1). The goals wording is deceptively short and simple; the path towards
its achievement is anything but. A healthy society is vitally important to American life:
health is a necessity at the individual level for quality of life, at the community level for
thriving cities and productive workforces, and at the national level for a growing
economy and prosperous citizenry. The problem is that America spends far more than
any industrialized nation in the world, yet ranks only mediocre in most important health
outcome measures, as detailed in Table 1.1 below. Ninety-seven percent of total health
care costs are related to medical care provided on an individual basis (Institute of
Medicine [IOM], 2012a). With costs continuing to increase, there is a consensus among
health care and policy leaders that reliance on the primary care model has become fiscally
unsustainable for the nation (IOM, 2012a). Further magnifying the challenges of a
healthy society is the fact that barriers to health are multifaceted, extensive and
interrelated.
Table 1.1 U.S. Health Rankings.
US Ranking out of total OECD Countries
Life Expectancy 26th of 34 Infant Mortality 30th of 34 (2008 data) Maternal Mortality 25th of 34 (2007 data)
Source: Organization for Economic Co-operation and Development [OECD], 2009.
1


Health Care and Public Health Systems
Ideas to improve the health care system abound, but what has become clear to
researchers and policy makers alike is that a systems approach is required. Health care in
America is a complex and multifarious set of systems, with layers of subsystems. There
are primary care and specialist physicians providing care to individuals, focused for the
most part, on treating existing medical symptoms and conditions. A bevy of insurance
companies, medical equipment and supply firms, pharmaceutical companies and billing
administrators are heavily invested in the provision of care, with their own biases and
objectives. Hospitals, some for-profit, some nonprofit, are part of the systems
infrastructure, along with government regulators, medical universities and other health
training facilities. Research and development, from both the private and public sectors,
continue to change the trajectories of the health care business models, clinical services,
future research priorities and medical education.
Over $2.6 trillion is spent each year on medical care in the health care system
described above, and yet that does not represent the whole picture (IOM, 2012b).
Ensuring a healthy nation clearly requires a system of health care that includes broad
access to providers and the provision of quality care, but it also requires efforts to prevent
disease and injury; efforts to promote healthy lifestyles; and the assurance of safe
conditions and environments in which to live and work (IOM, 2012b). Thus, there is the
additional layer of the public health system, which is focused on preventing health
conditions such as chronic disease, cancer, disability and injury, at the population level
(Detels, Beaglehole, Lansang, & Gulliford, 2011). Public health further increases the
2


complexity of health in America, bringing a whole new set of actors, funding streams,
objectives and level of focus.
The public health system in America is anchored by governmental public health
organizations, specifically, the 51 state public health agencies, 2,794 local public health
agencies, and 565 American Indian and Alaska Native tribal public health agencies
(National Association of County and City Health Officials [NACHO], 2010). These
organizations are charged with the task of providing 10 essential public health services
(EPHS), detailed in Key Terms and Definitions in this Chapter. However, these
governmental organizations do not have the capacity or funding to be able to provide the
EPHS alone. To fulfill their charge, they partner with organizations within all sectors,
serving as a catalyst for engaging multiple stakeholders to confront community health
problems (IOM, 2012a, p. 1-9.) Partners at the national level include federal agencies
such as The Health Resources and Services Administration (HRSA), Centers for Disease
Control (CDC), and National Institutes of Health (NIH); and national nonprofit
organizations such as the Institute of Medicine (IOM) and Robert Wood Johnson
Foundation. Partners at the state and community level include primary care providers,
universities and hospital systems, federally qualified health centers, community based
health centers, and mental health centers, among others. These partners highlight the
interwoven complexity and interdependency of the health care delivery system with the
public health system. Additional state and local partners may include organizations
outside the realm of primary care, such as business groups and faith based organizations.
Community participation is also vital to public health, and thus, local partners also often
3


include advisory boards, nonprofit and grassroots organizations, as well as families and
individuals (IOM, 2012a).
Traditionally perceived to be two separate systems, there is an increasing call for
primary care and public health to integrate because many organizations work within both
primary care and public health, sharing the common goal of improving health. The
integration of these two systems is seen as the crucial element to enhance the capacity of
both sectors to carry out their respective missions and link with other stakeholders to
catalyze a collaborative, intersectoral movement toward improved population health
(IOM, 2012a, p. S-l). It is hoped that better integration will enable both primary care and
public health to capitalize on their synergies, and better achieve an overall improvement
in heath together.
Statement of Research Problem
Given the complexity of tasks in assuring conditions for health, and the breadth of
organizations that are involved, it has been recognized by health leaders and policy
makers across the country that change at the system level is required to enable substantial
and sustainable improvements in the health of Americans (IOM, 2012a). A systems
approach, and an understanding of system level factors (and their impact on the delivery
of health care and public health services) is thus critical. Interorganizational relationships
make up an important system level factor: partnerships and collaboration are an integral
component of the public health system's organization and structure; an increasingly
common method of health and public health service delivery; and an important influence
for population health outcomes as an essential public health service unto themselves
(IOM, 2011).
4


Interorganizational Relationships
Interorganizational relationships are a core feature of the health and public
system's organization and structure for the fundamental reason that assuring conditions
for population health is beyond the scope and abilities of any one organization or
government agency (NACHO, 2012). Stakeholder groups and partnerships are the
foundation of the local public health infrastructure, and can include any individual,
group, or organization that affects, or can be affected by, actions to improve population
health. Through a range of relationships that foster the sharing of resources and
accountability for health improvement (National Public Health Performance Standards
Program [NPHPSP], 2007, p. 24), these interorganizational relationships enable public
health agencies to carry out their core functions and essential services.
Interorganizational relationships are clearly important in the health and public
health systems where public, private and nonprofit organizations work together to assure
the conditions for health; however, there is a significant lack of evidence in this area.
The Literature Review in Chapter III chronicles the lack of robust research focused on
effective and efficient interorganizational relationships in health and public health
systems. Further, the 2012 Public Health Services and Systems national research agenda,
created by research scientists and health and public health practitioners and leaders, has
identified research that focuses on interorganizational relationships in public health as an
important research priority (Consortium, 2012). The systems approach enables research
that emphasizes these important interrelations between and among the various actors
working in health care and prevention.
5


Theoretical Framework
In the field of systems theory, a system is defined as a set of elements standing in
interrelation among themselves and with the environment (National Cancer Institute
[NCI], 2007, p. 159). Thus, a systems approach requires investigation of the whole,
rather than just an actor or aspect within a system (Mabry, 2011; Mabry, Olster, Morgan,
& Abrams, 2008; Gillies, 1982).
Fundamentally, systems theory emphasizes that a system consists of a set of
actors and their relationships. Thus when the US health care and public health systems
are viewed through the lens of system theory, the relationships between and among the
many actors described above are as critical to the system as the actors to which they
belong. These relationships can have tremendous impact on the performance of the
system. To this end, there are significant opportunities to maximize and utilize these
relationships to leverage the resources, knowledge, information, and services to improve
the system, and ultimately, help to support a nation in which all people can live healthier
and more productive lives (United States Department of Health and Human Services
[HHS], 2012). The idea that a system is greater than the sum of its parts is at the heart of
this approach.
A systems framework, which builds upon system theorys holistic perspective and
assumptions of interdependencies, is thus a strong research approach to investigate
interorganizational relationships embedded in the health and public health systems. Such
an approach is further strengthened by theories of network structure, which aim to foster
an understanding of effective relationships among actors and stakeholders to improve
collaboration, information and resource flows and exchanges, and to help reduce
6


redundancies (Kilduff & Tsai, 2003). A novel methodology in a systems approach is
network analysis, which holds the potential for generating insight and facilitating
strategic management of relationships between actors and stakeholders within a system.
Many theories of network behavior underpin this type of systems approach, and the core
network concepts, as detailed in the methodology in Chapter IV, have broad acceptance
in systems research (NCI, 2007). The use of network analysis as a methodology enables
exploration of the interorganizational relationships that are so important to the complex
health and public health systems, and enables investigation of unanswered research
questions, such as those in this study.
Research Questions
Embracing a systems approach, and responding to both the PHSSR research
agenda (Scutchfield, Perez, Monroe & Howard, 2012) and the call to bridge primary care
and public health (IOM, 2012a), this study asks: Do community based stakeholder
networks operating within the same context exhibit commonalities in network-level
attributes and characteristics relative to their performance? The second research
question focuses upon organizational level variables, asking: Do the relative connectivity
and structural positions of member organizations affect stakeholder network
performance?
Study Design
In a mixed method, cross case research design, this study investigates five
stakeholder networks working in the context of early childhood intervention (El), a
subsystem that spans the health and public health systems. The El system at the national,
state and local levels is described in Chapter II.
7


The relational and network measures of the networks will be examined through
social network analysis (SNA), and analyzed with performance measures based on
external assessments, network self assessments, and early intervention services data.
Population: Focusing on the dynamic interactions of primary care providers and
public health agencies with other community based public and private organizations, the
population of this study includes five stakeholder networks working to ensure
standardized developmental screening, referral and follow up for children in their local
community in Colorado. These networks included stakeholders such as the local public
health agency, early intervention service provider, early childhood council and providers,
primary care providers, school districts and other organizations in the community with an
interest or concern with the local system of early intervention. These networks were
mobilized to overcome the segmented system of identification and delivery for early
intervention services.
Research Setting: Early intervention is a system of coordinated services that
promotes the growth and development of children during the critical early years of brain
development. The goal of early intervention is to assure that families who have children
with disabilities or developmental delays, ages birth to three, receive resources and
supports that assist them in maximizing their child's development. Research has shown
that skills and abilities can be fostered and developed during those early years of brain
development that may be unattainable if the learning process starts later in life
(Mackrides & Ryherd, 2011).
Utilizing systems theory as the general framework, and network theories to
examine specific measures, this investigation focuses on the interorganizational
8


relationships that are inherent in each stakeholder network within the El system that
spans the health care delivery system, the public health system, as well as human services
and education systems. This study investigates how different network attributes and
characteristics at the network level (referred to hereafter as whole-network) and the
individual organization/dyad level (referred to hereafter as organizational level) including
structure, connectivity, trust and value affect performance. Content analysis is used to
further investigate patterns that emerge from the network analysis.
Theoretical background: While systems theory focuses this investigation on the
interrelation among and between the multitude of organizations that are a part of the
health and public health systems to ensure childrens healthy development; network
theories provide an important frame of reference in their assertions that relationships are
not only productive resources, but that well structured networks can be more effective
than large inclusive networks (Moynihan, Provan & Lemaire, 2012; Burt, 1997). This has
a significant application in public health practice, where state and local public health
system performance is typically only assessed, in part, by counting the number of
stakeholders in partnerships (NPHPSP, 2007). Provan, Fish, and Sydows (2007)
typology of organizational network research provide a framework for analysis, while
Varda, Chandra, Stern & Luries (2008) core dimensions of connectivity; Karckhardts
(1992, 2010) strength of strong ties findings; Granovetters (1973, 1992) strength of weak
ties theory; and Burts (1995, 2000, 2001) brokerage research inform the
operationalization of network variables to analyze and explore patterns at the whole-
network and organization level across the five stakeholder networks.
9


Combining the quantitative methodology of social network analysis to measure
the social structure, with the exploratory nature of qualitative inquiry to explore emergent
patterns, this study aims to provide insight into the dimensions of partnership process and
performance. The studys findings provide new empirical evidence about how different
characteristics and attributes of partnership affect performance, adding to a much needed
knowledge base to guide the use, formation and management of interorganizational
relationships for effective health and public health systems.
Plan for Dissertation
This investigation is presented in six chapters, with the aim to provide insight into
the attributes and characteristics of stakeholder networks (working to improve the system
of early intervention), and their impact on network performance. Chapter II explores the
policy context of El nationally, and describes the segmented system of El. The chapter
describes how the El system is implemented at the state and local levels in Colorado. To
this end, it introduces the multiple legislated agencies, along with their respective goals,
from public health, human services and education, and describes the reasons behind the
call for greater integration of primary care providers. The chapter also details the specific
types of organizations that are working together in the five stakeholder networks in this
study.
Chapter III begins with the literature of systems theory, and offers an application
in the context of early intervention. The chapter then navigates the complex body of
social and organizational network theories. Demonstrating a lack of empirical evidence
that relates whole-network attributes to network performance, the exploratory approach
to the first research question is explained. The chapter then synthesizes the network
10


theories that inform the investigation into the second research question, as well as offers a
discussion on the methodologies and the assumptions contained within.
Chapter IV offers a detailed description of all aspects of the research design and
methods used in the study, including network analysis to measure the network attributes
and characteristics, and content analysis to explore the emergent findings from the cross
case analysis. The chapter describes the mixed-method paradigm utilized in this study, as
well as the specific approach and procedures for data collection. It also includes an
explanation of the epistemological grounding for the methods and addresses how they are
a strong fit for this line of inquiry. The chapter then offers a detailed description of the
dimensions of network performance, the dependant variable in this study, and the whole-
network variables and organizational level variables, the independent variables in this
study. The chapter concludes with the limitations of the study, along with the reliability
and validity of the approach.
Chapter V details the findings of the network analysis and identifies patterns that
exist across the five cases. The chapter is organized around the research questions, first
presenting results from the cross-case analysis of the whole-network variables; and then
presenting results from the cross-case analysis of the attributes and characteristics at the
organizational level. The chapter concludes with findings from the content analysis,
which was undertaken to better understand a distinction in organizational type that
emerged from the cross case analysis.
Chapter VI concludes the study with a discussion on the research findings. The
first part of this discussion is organized around the research questions, and offers a
summary of the findings, interpretation of the data, and conclusions drawn from the
11


information. A discussion about the contribution of the research to organizational
network theory is offered, as well as the implications for health and public health
systems. Four conclusions summarize the contributions to both knowledge and practice.
The chapter ends with recommendations for future research.
Key Definitions and Terms
System-. A system is a collection of independent elements that are interrelated
among themselves and within their environment, and organized in a meaningful way to
accomplish an overall goal (Hayajneh, 2007).
Systems approach. An approach to research or analysis that uses systems theory
and methods in an organized framework to investigate systems. The National Cancer
Institutes Initiative on the Study and Implementation of Systems (NCI, 2007)
synthesizes the four key systems approaches as:
1. Systems organizing to understand and foster the development of participatory,
complex, and adaptive collaborative systems; ensure their effective facilitation
and management; and encourage productive system action and learning.
2. System dynamics to understand and model the complex dynamic interactions
involved in the system.
3. System networks to understand and analyze effective collaborative relationships
among stakeholders, improve collaboration strategies, and help reduce duplication
of effort.
4. Systems knowledge to develop and manage the knowledge infrastructure
required for effective dissemination and evolution of scientifically credible,
12


evidence-based practices, together with an effective strategy to package, deliver,
and maintain this knowledge (p. 2).
Systems methods are tools, techniques or procedures to enable investigation of systems.
Network analysis, system dynamics modeling, and structured conceptualization are all
systems methods (NCI, 2007).
Primary Care: Primary care in America is defined by the Institute of Medicine
(IOM) (1996, p. 1) as the provision of integrated, accessible health care services by
clinicians who are accountable for addressing a large majority of personal health care
needs, developing a sustained partnership with patients, and practicing in the context of
family and community. Most primary care is delivered through the private sector, but it
can also be delivered through government agencies, including the Veterans Health
Administration (VHA) hospitals and Health Resources and Services Administration
(HRSA)s safety net clinics. Regardless of whether it is private or public, primary care
physicians often provide a point of first contact for health related issues, and aim to offer
comprehensive and coordinated care (Starfield & Horder, 2007). The medical home
model of primary care encompasses the characteristics of care that are considered
essential for all children: accessible, continuous, comprehensive, family centered,
coordinated, compassionate, and culturally effective (Strickland, Jones, Ghandour,
Kogan, & Newacheck, 2011, p. 605).
Public Health. The IOM defines public healths fundamental charge as fulfilling
societys interest in assuring conditions in which people can be healthy (IOM, 1988, p.
140). Public health encompasses a diverse group of public and private stakeholders
13


(including the health care delivery system) working in a variety of ways to contribute to
the health of society.
Essential Public Health Services: The ten Essential Public Health Services
(EPHS) provide a more specific definition of public health and a framework for the local
public health systems. Each of the states public health departments and local health
agencies are charged with carrying out the EPHS. The Center for Disease Control and
Prevention (2012) details the EPHS:
1. Monitor health status to identify and solve community health problems.
2. Diagnose and investigate health problems and health hazards in the
community.
3. Inform, educate, and empower people about health issues.
4. Mobilize community partnerships and action to identify and solve health
problems.
5. Develop policies and plans that support individual and community health
efforts.
6. Enforce laws and regulations that protect health and ensure safety.
7. Link people to needed personal health services and assure the provision of
health care when otherwise unavailable.
8. Assure competent public and personal health care workforce.
9. Evaluate effectiveness, accessibility, and quality of personal and population-
based health services.
10. Research for new insights and innovative solutions to health problems (para.
3).
14


Significance
Building a health care system that meets the needs of all Americans is an
ambitious goal for the government, and requires both comprehensive, well-informed
policy, as well as successful implementation through effective systems.
Interorganizational relationships are fundamental in health and public health systems
where public, private and nonprofit organizations work together to assure the conditions
for population health. Yet given the significant reliance on these relationships, there is
still much to learn. Uncertainty abounds about when organizational networks are an
effective or efficient choice, and how to build and foster effective and efficient
interorganizational relationships. It is important that health practitioners and leaders are
armed with the knowledge to prompt strategic and thoughtful discussion to foster
stronger inclusion, involvement and commitment where it most makes sense and will
have the greatest impact. An understanding of interorganizational relationships is thus
critical in this endeavor, and has twice been highlighted as an area of needed research in
the last decade (Scutchfield et al., 2012).
The gap in knowledge may exist because traditional research models are
challenged by the large and complicated array of actors and variables within health and
public health systems, and are often not able to take into account that fact that the
systems are comprised of interrelated actors, and are interrelated themselves. Further, the
systems are dynamic: they are always changing as feedback from one element of the
system changes the course for another. A majority of the research into health care
delivery in the twentieth century has been through reductionism, the process of
attempting to understand a problem by first deciphering its components (NCI, 2007, p.
15


13; Barabasi, 2002). These traditional approaches have left important research questions
unanswered and tremendous gaps in our knowledge of how to improve the health and
public health systems in America. Given this, a systems approach, which guides this
study, has gained more attention and traction in the research and practitioner arenas as a
way to examine and explore the interrelated elements of the systems, and ultimately learn
how to better improve and support the health of Americans.
This investigation aims to provide stakeholders the knowledge to begin
positioning their stakeholder networks to most effectively achieve their goals and
positively influence health outcomes, while maximizing the return on their investment
within the networks. Ultimately, stakeholder networks need to be effective in achieving
their goals in order to be useful as a tool for policy implementation, and the better we
understand the dimensions of interorganizational relationships that affect their success,
and how, the more equipped we are to harness and leverage the strengths of these
relationships to improve health outcomes in America.
16


CHAPTER II
RESEARCH CONTEXT: SYSTEM OF EARLY INTERVETION
Background
Americas aim to be a country where all citizens live long, healthy lives is a
considerable challenge, with a complex set of systems in place to carry out the mission
(HHS, 2012). Inherent in such an endeavor is the important goal of meeting children's
health needs, especially for those children who are born with disabilities or those who
experience delays in development during their early childhood. The literature shows that
12-16 percent of children in America have at least one developmental delay (Mackrides
& Ryherd, 2011). Early intervention (El) systems exist to provide coordinated supports
necessary for these children to reach their full potential. El is a system of coordinated
services that promotes the child's growth and development and supports families during
the critical early years of early brain development. The El system is especially complex,
given that it is legislated at the federal level but implemented at the state and local levels.
Further, the El system extends across the public health system, systems of education
and/or human services, depending upon the state, all the while being called to integrate
with part of the health care delivery system. Set at the nexus of primary care and public
health integration, this study is focused on this multifaceted subsystem for childhood
development.
Early Intervention Services
The purpose of El is to ensure that families who have children with disabilities or
developmental delays, ages birth to three, receive resources and support that assist them
in maximizing their child's development. Early intervention services are focused on the
17


development that babies typically experience during the first three years of life,
including: cognitive development (such as thinking, learning, solving problems); physical
skills (reaching, rolling, crawling, and walking); communication; and social development
(National Dissemination Center for People with Disabilities [NICHY], 2012).
Early intervention services provide vital support so that children with
developmental needs can thrive and grow. Services are often provided by physical
therapists (PT), occupational therapists (OT), and speech and language pathologists,
audiologists, psychologists, early childhood educators or nurses. These therapists work
with the children to develop neural pathways, muscles, skills, and knowledge that are not
developing in a typical manner, and provide the foundation for language, reasoning,
problem solving, behavior, and emotional health. Therapists also provide support services
to the families such as family education and counseling, parent support groups, special
instruction, or assistive technology devices and services (NICHY, 2012).
There is a substantial body of research that demonstrates the power and necessity
of providing intervention services to children with disabilities and developmental delays
at young ages. Research has shown that early intervention services may:
Improve childrens gains in developmental, social, and educational area;
Reduce future costs of services and health care needs (such as special education,
therapies or rehabilitation) for both families and government systems;
Reduce stress and feelings of frustration for children and families;
Help children to reach their maximum potential and become productive,
independent individuals (Bright Tots, 2012).
18


Although there is still much to be learned about the services that work best for different
children and families, the combined body of existing research clearly underscored the
need to intervene early to enable children with delays and disabilities to reach their full
potential (Hebbeler, Spiker, Bailey, Scarborough, Mallik, Simeonsson, et al., 2007, p. 1-
1).
The benefits of El services are felt beyond the children and families who receive
them. Research also shows that such services can reduce the prevalence of
developmental and behavioral disorders as children get older. Further, researchers Hess
and Van Landeghem (2005, p. 3) have found that early intervention efforts for children
and their families are not only beneficial and cost-effective to the health system, but can
also reduce the prevalence or necessity for more costly interventions and outcomes such
as welfare dependency and juvenile detention. If children do not receive needed
services and support, they are not only less likely to reach their developmental potential;
there may also be long-term consequences for the health, education, child welfare, and
justice systems.
The importance of early intervention services for infants and toddlers with
disabilities is well documented in the literature (Jones & Ziglar 2002; Haskins, 1989);
that is not the investigation here. When children receive El services, there seems to be
little doubt in the research of the positive benefits for the child directly, as well as
indirectly to government funded health, education, child welfare, and juvenile justice
systems. The problem is that the implementation of the policy is through a segmented
system of care. That identification and services for El are divided between public health,
19


human services, education, and primary care, and resulted in children not always
receiving the care they need.
Grounded in the health and child development research, three trends at the
national levels created the basis of the early intervention system. Title V of the Social
Security Act legislates the public health system to support El, while Part C of the
Individuals with Disability Act (IDEA) legislates the systems of education and human
services to provide El services; all the while, medical associations such as American
Academy of Pediatrics urge pediatricians and primary care providers to more
systematically provide screening and referrals for young children who may need El
services. The El system thus spans separate and distinct systems: public health,
education, human services, and health care (primary care), each with its own unique role
to play. Figure II. 1 illustrates the three silos at the national, state and local levels that
exist due to the way the services have been legislated. Given this, inter organizational
relationships are critical in EEs segmented system of services; not only to enable
organizations within each system to work together efficiently and effectively, but also to
bridge the separate systems. These interorganizational relationships may provide the
leverage to improve the El system for the benefit of children, and act as a model for the
health care and public health systems as a whole.
20


Social Security Act, Title V
Individuals with Disabilities Act,
PartC
National health professional
associations: recomendations
and guidance
State Health Departments
(CDPHE in Colorado)
State Departments of Education of
Human Services
State chapters and
medical societies
(CDHS in Colorado)
Local Health Agencies
(55 in Colorado)
Local Human Services or
Education Agencies
(20 CCBs in Colorado)
Individual physicians and
helath care providers
Figure II.1 Systems That Make Up the Early Intervention System at the National,
State and Local Levels.
Early Intervention and the Public Health System
National Policy Context: Title V Maternal and Child Health Program of the Social
Security Act
Legislated in 1935 as a part of the Social Security Act, the Title V Maternal and
Child Health Program lays a foundation in public health for ensuring the health of
women, children and youth (including those with disabilities or special health care needs)
in America. Among many things, Title V aims to:
Provide and ensure access to preventive and child care services as well as
rehabilitative services for certain children;
Increase the number of children receiving health assessments and follow-up
diagnostic and treatment services;
Implement family-centered, community-based, systems of coordinated care
for children with special healthcare needs (MCHB, 2012a, para. 2).
21


Title V established the Maternal and Child Health Bureau (MCHB) at the federal
level, which aims to provide leadership, in partnership with key stakeholders, to
improve the physical and mental health, safety and well-being of the maternal and child
health (MCH) population which includes all of the nations women, infants, children,
adolescents, and their families, including fathers and children with special health care
needs (MCHB, 2012b, para. 2). The program plays a critical role in quality oversight,
coordination (between and among programs and agencies), and overall system capacity
building (MCHB, 2012b).
State Policy Implementation
Maternal and Child Health Services Block Grants (Title V Block grants) are
awarded to State Title V programs, which are usually located within a maternal and child
health division of state health departments. Each year, these programs apply for: State
Formula Block Grants, Special Projects of Regional and National Significance
(SPRANS), and/or Community Integrated Service Systems (CISS) projects (MCHB,
2012).
Such implementation allows states and jurisdictions tremendous discretion in how
they use of their Title V funds, with the aim to tailor their programs to meet the unique
needs and challenges in their states. Each state must identify priorities that address the
needs of their MCH population, and then guide the use of the Maternal and Child Health
Block Grant funds to those ends. To do so requires a partnership arrangement between
Federal, State and local entities, and is guided by a state-wide needs assessment
conducted in collaboration with other organizations in the state and local communities
every five years. Partnerships are thus a required element of the grant. While discretion
22


exists, each state must use its Title V funds to focus exclusively on the maternal and child
health population for both infrastructure and service costs. Further, the funding requires
a match of 75% (meaning that $1 of Federal Title V funding must be matched by at least
$.75 of state and/or local funding).
At least 30 percent of federal Title V funds are earmarked for preventive and
primary care services for children; and at least 30 percent are earmarked for services for
children with special health care needs, enabling support for early intervention services as
necessary. The funds may also serve as the payer of last resort for services to children
without coverage from another program (Catalog of Federal Domestic Assistance
[CFDA], MCH Block Grants, Number 93.994, 2012).
A total of 59 states and jurisdictions (including the District of Columbia and U.S.
Territories) received Title V Maternal and Child Health Block Grant funding in fiscal
year 2011, ranging from $145,927 $41,389,219 per state and jurisdiction, as detailed in
Table II. 1. The required match results in almost $1 billion being available annually for
maternal and child health programs at the state and local levels. In fiscal year 2011, State
Title V programs served over 39 million individuals. Among the individuals served were
2.5 million pregnant women, 4.1 million infants, 27.6 million children, and 1.9 million
children with special health care needs (MCHB, 2012a; CFDA, MCH Block Grants,
Number 93.994, 2012).
23


Table II. 1 Number of Individuals Served and Annual Appropriations for Title V of
Social Security Act.
Federal Fiscal Year 2011
Number of States and Jurisdictions 59
Number of Individuals Served 39 million
Appropriations $546,792 million
Source: Catalog of Federal Domestic Assistance, 2012.
Implementation in Colorado
The Colorado Department of Public Health and Environments Prevention
Services receive and distribute the Title V block grants to the 55 local public health
agencies implementing locally in Colorados 64 counties. The local public health agency
staff members are charged with implementing the local action plans in their communities
(Colorado Department of Public Health and Environment [CDPHE], 2012a). For
example, the Division of Maternal and Child Health (MCH) at CDPHE is working to
improve developmental standardized screening and referral rates for all children,
newborn to age 5, and listed it as a 2011-2015 priority (CDPHE, 2012a). The Children
and Youth Branch (CYB) at CDPHE also receives funding from local sources to help
integrate health into local early childhood systems-building efforts (CDPHE, 2012b).
Title V was a great start, but as more research citing the significant benefits of
early intervention was published in the 1970s and 1980s, along with a recognized need
for improved accessibility to education for individuals with disabilities, an El system of
services was legislated in Part C of Individuals with Disabilities Act (IDEA).
24


Early Intervention and the Education and Human Services Systems
National Policy Context
The overwhelmingly positive findings in the research for early intervention led to
legislation mandating El services in 1986. The Program for Infants and Toddlers with
Disabilities (Part C of the Individuals with Disabilities Education Act [IDEA]) legislates
state-based programs of early intervention services for children aged 0-3 who have
developmental delays and disabilities. This federal law was grounded in a body of
existing research [that] clearly underscored the need to intervene early to enable children
with delays and disabilities to reach their full potential (Hebbeler, et al 2007, p. 1-1).
The Education for all Handicapped Children Act of 1975 forged the path towards
what would eventually become IDEA. In this act, federal funded public schools were
required to provide equal access to education for children with disabilities. IDEA thus
focuses on the educational needs of children with disabilities from birth to age 21, which
is why El services legislated in the Act are provided through education and social
services agencies, rather than public health or primary care. The program emphasizes the
need for cross-agency coordination, again underscoring the need to better understand and
improve interorganizational relationships.
State Policy Implementation
IDEAS Part C provides federal grants to states to implement early intervention
services such as speech therapy, physiotherapy and occupational therapy for eligible
children from birth to age three. Again, states have considerable flexibility in use of this
funding, which helps the states better meet their own unique needs, but also results in
large variations in the proportion of children enrolled, ranging from one percent to seven
25


percent of early childhood populations in each state (The Early Childhood Technical
Assistance Center [ECTAC], 2008; Russ, Garro, & Halfon, 2010). In fiscal year 2011,
nearly 336,895 children under age 3 were served by IDEA Part C programs, as detailed in
Table II. 2 (United States Department of Education, 2012).
Table II.2 Number of Children Served and Annual Appropriations for Part C of
IDEA.
Federal Fiscal Year 2012
Child Count Year (Fall) 2011
Number of Children Served 336,895
Percent of Population 2.79%
Appropriations $442,710 million
Source: United States Department of Education, 2012.
Some states have also been able to expand services to children who are at
risk of disabilities or developmental delays, many of whom come from
low income families (Russ, Garro, & Halfon, 2010).
Implementation in Colorado
The Colorado Department of Human Services, Division for Developmental
Disabilities administers the Early Intervention Colorado Program and contracts with
twenty Community Centered Boards (CCBs) statewide to provide these early intervention
supports and services to children (newborn to age three), and their families (CDHS,
2012). The CCBs are responsible for determining eligibility, providing case management
and providing or contracting for services and supports (CHDS, 2012). The CCBs also
serve youth and adults, outside the focus of the El system.
26


Limitation in Early Intervention System
Despite almost three decades of implementation of PART C of IDEA, evidence
exists that not all children are receiving the care they need. One of the most significant
barriers lies in the challenge in identifying children who have developmental delays
during those critical early years of brain development. Kaye, May and Abrams (2006)
found:
many young children are not identified with developmental problems until school
entry or until they demonstrate school failure... [and that] although more than 95
percent of young children see a child health care clinician in the first three years
of life, most of these clinicians are missing opportunities to detect developmental
problems, counsel parents of young children about developmental issues, or refer
children to needed services in the community, (p. 4).
These findings are supported by the National Survey of Early Childhood Health, which
found that 94 percent of children had parents who were not getting the guidance or
education they need as it relates to the screening of their child (Bethell, Reuland, Halfon,
& Schor, 2004). Dr. Neal Halfon agrees, stating at the Surgeon General's Conference on
Children's Mental Health in 2000 that a majority of problems go unrecognized, and most
children do not receive treatment early in their life unless the problems are severe (HHS,
2005 p. 21). Developmental delays are also especially prevalent in young children in low
income families, and researchers have found that these children are significantly under-
detected (Kaye et al., 2006).
In addition to a segmented El system, America consistently ranks among the
lowest in cross-country comparisons of child well-being (Wise & Blair, 2007). This is
27


made even more important given the body of evidence that demonstrates the link between
adverse early life experiences and development, and poorer health or chronic diseases
later in life (Kuh & Ben-Shlomo, 2004; Shonkoff & Phillips, 2000). Researchers have
also demonstrated the powerful links between mental health in childhood and later
economic well-being (Currie, Stabile, Manivong & Roos, 2008).
Why are some children not receiving the care they need? The systems approach
introduced in Chapter I illuminates the segmentation in the El system of care. From this
perspective, the root of the problem may be that the delivery for early intervention
services, which is provided through education and human services agencies, is
disconnected with the primary care system, where pediatricians routinely see children
birth to three for well-child visits and checkups. Neither education nor human services
would typically have access to children ages birth-three, unless they are referred to those
systems (Hebbeler et al., 2007). Identification and referral takes place in other systems,
specifically the health care system, and to some degree, in early childhood education.
Early intervention agencies may be well run organizations, but they alone cannot ensure
that all children who are in need of services receive them: these agencies are set up to
react to a referral for services. Further, the lead agency that administers the program
varies among states, as do the funding mechanisms and the collection of programs that
actually provide the services. This recognition has led to a call to better integrate the
primary care providers in the health care delivery system into the El system.
28


Health Care System
National Context
The American Academy of Pediatrics (AAP) is a professional membership
organization of over 60,000 primary care pediatricians, pediatric medical sub-specialists
and pediatric surgical specialists dedicated to the health, safety, and well being of
infants, children, adolescents and young adults (American Academy of Pediatrics
[AAP], 2012, Membership, para. 1). The American Academy of Family Physicians
(AAFP) represents over 100,000 family physicians, family medicine residents, and
medical students and aims to transform health care to achieve optimal health for
everyone (American Academy of Family Physicians [AAFP], 2012, Vision Statement).
These two groups of physicians make up a considerable part of the health care system in
America.
The AAP (2012, para. 1) aims to attain optimal physical, mental, and social
health and well-being for all infants, children, adolescents and young adults, and as
such, makes recommendations that often become the basis of pediatric health care. The
organization published a report in 2001 highlighting consensus in the field that pediatric
clinicians have both the opportunity and expertise to identify children in need of El
services to support their development, stating that there is growing consensus on the
important role that primary care providers, who see the child on a regular basis and can
thus assess development over time, can play in recognizing potential developmental
problems, including social and emotional development problems (AAP Committee on
Children with Disabilities, 2001, pp. 192-195). The report goes on to assert that
pediatricians, with both the opportunity and expertise, are ideal candidates for conducting
29


the ongoing surveillance needed to identify developmental problems (AAP Committee on
Children with Disabilities, 2001).
Further, the American Academy of Pediatrics (AAP) also developed the medical
home concept, which is becoming the standard for provision of high-quality
comprehensive health care (Long, Bauchner, Sege, Cabral, & Garg, 2012). The medical
home concept is defined as "model of primary care that is accessible, continuous,
comprehensive, family-centered, coordinated, compassionate, and culturally effective"
(Strickland, et al., 2011, p. 605). The model requires that children have a consistent
health care provider (a physician or nurse) who works with the family to best meet the
needs of the child, with access to referrals and care coordination as needed (Strickland, et
al., 2011).
Implementation in Colorado
Many of the professional associations such as AAP and AAFP have active state
chapters that support individual pediatricians and family medicine physicians practicing
in Colorado, whether they are based in private partnerships or solo practices throughout
the state, or work for federally funded VA clinics or HRSA safety net clinics. The
Colorado Medical Society and other state based organizations also play a key role in the
primary care system in Colorado, as do physician assistant and nurse practitioner
organizations.
Segmented Early Intervention System
As mentioned above, there are separate steps in the El process, each taking part in
a different system. First and foremost is identification and referral, represented as red in
Figure II.2, which often takes place within a primary care setting, when initiated by a
30


physician or referred to a physician, or within public health through a safety net clinic or
home visitation program. There is then the eligibility process that is facilitated by the
CCBs on behalf of human services. This may also include screening, if not conducted by
a physician beforehand. If deemed eligible, the CCB provides the El services,
represented in purple in the figure below, or contracts therapists to provide the services.
On the childs fourth birthday, they must transition out of the human services system to
the education system, represented by the blue arrow below.
Identification Provision
and Referral Elgibility of El Education
Services Services
Figure II.2 Steps in Early Intervention Process, Cross Listed with System.
Mackrides and Ryherd (2011, p. 544) found that while 12 16 percent of children
in the United States have a developmental delay, as many as one-half of affected
children will not be identified by the time they enter kindergarten. If developmental
delays are not detected until they are five, opportunities for developmental gains through
early intervention are lost. Further, without well defined inter-agency and inter-system
processes, there are opportunities for a child to be accidentally dropped between the
systems.
Finally, a system perspective highlights an extremely important facet of the El
system, that no one agency or system can serve the child alone: supporting a child with a
31


disability or developmental delay to reach their full health and developmental potential
requires successful action in all parts of the systems.
Identification and Screening: Health care system involvement
While some families may know that early intervention services will be essential in
helping their child grow and develop from the time of birth (if the child is born with a
specific disability that is diagnosed at the hospital, for example), other children may have
a relatively routine birth, and then develop more slowly or quite differently than other
children. These children require someone to identify their developmental delay(s) and
refer them for El services.
There is evidence in the literature that physicians who use an objective screening
tool will more effectively identify children who may be at risk for, or have, a
developmental delay than physicians who do not use such a tool (Rydz, Shevell,
Majnemer & Oskoui, 2005, p. 4). Several studies also indicate that using a developmental
screening tool improves the accuracy with which children are identified when compared
with decisions based only on clinical judgment (Rydz et al., 2005, p. 8). However, there
are indications that pediatricians and family medicine physicians do not regularly use
standardized tools for reasons that include a lack of time to administer screens during
health visits; inadequate compensation; lack of training in the use of specific tools; and
lack of, or perceived lack of, assessment and treatment resources (Sices, 2007, p. 9;
HHS, 2000). Although most physicians report using informal developmental checklists
as part of their overall strategy of care, a literature review by National Academy of State
Health Policy found indications that a low percentage of children in need of support for
32


their healthy development are identified, even by physicians (Sices, 2007; Rydz, et al.,
2005).
The literature on clinical recommendations for screening in primary care is
inconsistent and often insufficient to direct the primary care physician. In addition,
multiple barriers exist, including time limited appointments and lack of resources, which
often prevent physicians from performing initial screening and completing additional
evaluation and referrals (Mackrides & Ryherd, 2011; Sices, 2007).
If a child starts at elementary school with an undiagnosed disability or
developmental delay, the child stays in the education system to receive special education
services; however, a critical period of brain development has already passed, and the
child may not be able to make the therapeutic or clinical advances that might have been
possible through early intervention (NICHY, 2012).
Referral to Early Intervention
When viewing early intervention through a systems approach, it becomes clear
that improving the screening and referral of children with developmental delays will help
to overcome some of the segmentation in the system. However, even if screening
improves through actions of the pediatricians and primary care providers, it will do little
good for the child if there is no coordination between the health care system and human
services and education systems. Further to this point, research has shown that providing
families and clinicians access to resources for assessment and treatment are also critical
(Kaye, 2006). Some researchers found that pediatric clinicians [are] reluctant to adopt
(or continue) using a screening tool unless they [are] confident that the children they
33


identified as potentially needing further care would receive appropriate care (Kaye, 2006,
p. 26).
Provision of Early Intervention Services
As mentioned, the Colorado Department of Human Services, Division for
Developmental Disabilities administers the Early Intervention Colorado Program in
Colorado. It provides early intervention supports and services through contracts with
twenty Community Centered Boards (CCBs) across the state, which in turn, provide the
services to infants, toddlers, and their families within their communities. The El system
in Colorado is visually depicted in Appendix 1.
It is clear that the El system in Colorado crosses many systems: public health has
worked for years to ensure care through maternal and child health programs and
initiatives, as funding from Title V programs at the national level is provided to CDPHE,
who in turn, provides it to local health agencies throughout the state. With the creation of
IDEA, the services for El are formally legislated and funded, but administered within
multiple systems: in Colorado, they are housed within human services and education
systems. Yet, the important time for El services is birth to age 3, a time before children
start with the education system. Human services may not have involvement in a childs
life at all, unless referred. And thus the El system is also dependent upon the health care
system, with a significant reliance on primary care providers to screen and identify
children with developmental delays. Further, there is a new acknowledgement that early
childhood directors and teachers can also help to indentify children with developmental
delays.
34


Despite Americas relative prosperity in the world, and the substantial health care
expenditures for seniors (via Medicare), the government makes only a relatively modest
investment in early childhood development (Bennett, 2008a & Bennett, 2008b). Through
the health and public health systems that exist, and with coordination with human and
educational systems, America must be able to leverage funding and resources to be as
effective as possible for our young children. As discussed, interorganizational
relationships are a critical element in this endeavor.
Implementation Challenges
In summary, systematic reviews of the research on early intervention have shown
that El services can enhance physical, emotional, and psychological growth for children.
Such services have also demonstrated positive short-term and long-term effects for the
child, family and government systems (Barnett, 1995; Jones & Zigler, 2002).
The literature also illustrates the highly segmented structure in the delivery of
early intervention services, which are provided through education and human service
sectors, while recent consensus emphasizes the parallel role of screening and
identification as part of primary medical care (King & Glascoe, 2003; Nelson). This
heightened awareness of the role of the primary health care system coincides with the
popularity of the growing research on early brain development, which places child
development in a biomedical context (Berry, Kutz, Langner & Budetti, 2008, p. 481).
The recognition that the health care system (including public healths safety net clinics)
may represent a more logical first point of contact for infants and young children (than
education and human services) has gained considerable strength, and further emphasizes
35


the need to strengthen and leverage the ties between primary care and public health, and
between primary care/public health and education and human services systems.
Recognizing these significant issues, and to stimulate innovative models for
integrating child development and medical care, the Commonwealth Fund, a private
foundation dedicated to improving the health care system, sponsors the Assuring Better
Child Health and Development (ABCD) initiative. The initiative assists states in
improving the delivery of early child development services by strengthening primary
health care services and systems that support the healthy development of children from
newborn to age three (Berry et al., 2008). The ABCD program focuses on building
system capacity to increase and enhance the delivery of child development services for
low-income children through the health care sector (Berry et al., 2008, p. 32).
Sponsored by the Commonwealth Fund and administered by the National Academy for
State Health Policy (NASHP), the program has been operating since 2000 in various
phases, helping a total of 27 states create models of service delivery and financing
through a laboratory for program development and innovation (National Academy for
State Health Policy [NASHP], 2012, para. 1).
Assuring Better Child health Development (ABCD] in Colorado
ABCD Colorados vision is that Colorados children reach their maximum
developmental potential (ABCD, 2012, para. 1). This guides its mission to encourage
the use of standardized developmental screening tools in health care settings across
Colorado to facilitate early identification and referral (ABCD, 2012, para. 2).
Colorados ABCD project started as a pilot project in Arapahoe County in 2005. The
program then received funding to expand outreach to the entire state two years later, and
36


a state team was developed to guide the efforts. The team includes pediatric primary
health care providers, parents, and leaders from state organizations and health care
systems. ABCD Colorado has worked with primary health care providers in over 40 of
Colorados 64 counties (ABCD, 2012) since 2005.
ABCD Colorado aims to support health care physicians by promoting early
identification of young children with potential developmental delays and referral to
appropriate community services and resources. It assists primary care practices in
implementing standardized developmental screening in an efficient and practical way;
and aims to help a physician practices build and strengthen relationships with early
intervention resources and services in their community (ABCD, 2012b). It also provides
training to other professionals who work with children during the critical ages from
newborn to age three, including child care and early education providers. This might
include training on developmental milestones, standardized developmental screening
tools, and appropriate referral processes (ABCD, 2012c).
ABCD Colorado supported community-driven stakeholder networks to form in
most of Colorados 64 counties. Five of these networks are being investigated in this
study. The stakeholder networks are working to ensure standardized developmental
screening, referral, and follow up for children in their own local community in Colorado.
While each group varies in nature due to the communitys own unique needs, most of the
stakeholder networks in this study include the following organizations and agencies, each
described in greater detail in Chapter IV:
Local public health department or agency (PH)
37


Local Community Centered Board, the El service provider contracted by the
Colorado department of Humans services (El)
Local school district or Board of Cooperative Education (ED)
Local pediatrician or primary care provider (PC)
Local Early Childhood Council (ECC)
Local early childhood centers/preschool (EC)
Local behavioral health center (BH)
Conclusion
In an effort to ensure that all children are able to reach their maximum
development, the stakeholder networks operating within the El system provide an
excellent set of cases to investigate, as the local implementation must be collaborative:
many organizations must work together with a number of government agencies, private
health care practices, and community based non-profits to overcome the limitations of a
segmented system. The ABCD program has a grant funded state team to guide the
program, and with its leadership, it supports community-driven, autonomous stakeholder
networks working in most of the states counties to ensure that every child in Colorado
has the opportunity to reach their maximum developmental potential (ABCD, 2012). This
investigation will examine and analyze multiple variables of process, structure and
connectivity shown to be related to a stakeholders ability to achieve its goals through a
comprehensive analysis of five stakeholder networks working to improve the El system
in Colorado. The segmentation and limitations of the El system lends to the importance
of being able to leverage interorganizational relationships to improve system capacity and
have a positive over influence of the health of Americas young children.
38


CHAPTER III
LITERATURE REVIEW
Introduction
The legislative context and policy implementation of early intervention, as
described in the preceding chapter, leads to a system where no single organization can
ensure that children receive necessary early intervention services. An effective and
efficient system of El services in Colorado requires primary care providers and early
childhood educators to identify and refer children who may benefit and be eligible for El
services. It also requires that the system incorporate feedback and follow through on
referrals, a focus of public health. Further, it requires human services and education to
coordinate an effective transition every time a child turns four and moves from one
system to the next for continued services. An effective El system thus require a number
of organizations and agencies with varied goals, resources, knowledge, and information
to work together to ensure childrens healthy development; highlighting the need for
effective interoganizational relationships.
Guided by Theory
This study examines the questions of who works with whom in stakeholder
networks working to improve their local El systems, and how the attributes and
characteristics of the interorganizational relationships within stakeholder networks impact
performance. Grounded in a systems approach, this study accepts the fundamental
assumption of systems theory that a system is greater than the sum of its parts, and
requires investigation of the whole, rather than specific aspects of a system, issue or
problem (Von Bertalanffy, 1972). Thus, this investigation focuses on the interrelation of
39


organizations within the system of El, which exists within the larger context of the health
and public health systems. More specifically, this approach utilizes the concepts in
network theories and the applications of network analysis to ultimately examine how
different attributes and characteristics of interorganizational relationships affect the
performance of stakeholder networks.
While the theoretical lens of a systems approach offers great value in its holistic
perspective of illuminating how and why a system is more than the sum of its parts,
systems theory requires additional theoretical support in a mixed, method cross case
research design, such as this one (NCI, 2007). Network theories, which focus on
relationships between and among actors, are powerful in their mathematical examination
of relational structure and ties. Network research offers a collection of theories and
frameworks that correlate the processes through which relationships can exert influence
on outcomes at the individual, organizational, or network level (Gulati, Nohria & Zaheer
et al., 2000). Although missing an overarching theory, network research fundamentally
proves that social structure matters (Ahuja, Soda & Zaheer, 2012). Network research in
the interorganizational context (hereafter referred to as the network approach), is
increasingly popular in health and public health research, and provides strength in
undertaking and explaining not only the interaction, connection and structure of
organizations partnering in health and public health, but also their affect on performance
and health outcomes (Wholey, Gregg & Moscovice, 2009; Mabry, 2011).
Combined, these frameworks offer a powerful theoretical and methodological
base for this study. By embedding the network approach within the overarching system
theory, one can leverage the strengths of each framework, while minimizing their
40


respective weaknesses. To this end, the literature review will reintroduce the research
questions, follow with a discussion of systems theory and then explore the network
approach.
Research Questions
With an emphasis on interorganizational relationships in health and public health
systems, and the embedded stakeholder networks structures and performance, this study
aims to answer the following research questions: do community based stakeholder
networks operating within the same context exhibit commonalities in whole- network
level attributes and characteristics relative to their performance? The second research
question focuses upon organizational level attributes and characteristics, asking: do the
relative connectivity and structural positions of member organizations affect stakeholder
network performance? As described, this study examines stakeholder networks working
in the context of early intervention, a system that spans health and public health systems,
as well as human services and education systems.
Systems Theory
Overarching Theoretical Framework
The foundation for systems theory has a long and storied history, traceable back
to Aristotles hypothesis that formal nature is of more importance than material nature,
known in modern times as the simple principle that the whole is greater than the sum of
its parts (NCI, 2007). General systems theory was developed by Ludwig von Bertalanffy
(1972), a biologist who crafted the theory to make sense of system characteristics such as
wholeness, differentiation, and order (Gillies, 1982). These roots of systems theory have
41


informed new theories including chaos theory and nonlinear dynamics, relational
mathematics, game theory, system dynamics; and network theories.
Systems theory frames the world as a complex system, composed of subsystems
that interact with each other. Each subsystem has clearly defined boundaries and
coherent dynamics (Von Bertalanffy, 1972). The theory can be used to understand the El
system, including its structures, processes and outcomes, and their interactions with the
health care and public health systems. Systems theory is thus used in this investigation as
an overarching theoretical framework to indentify the elements of the El system; the
relationships between these elements; and the systems boundaries and goals. Network
theories are then used to identify and operationalize the attributes and characteristics of
the stakeholder networks that include the actors within the system, working to improve
the system.
System Characteristics. According to systems theory, most systems have the
following common characteristics:
The elements of a system comprise a whole that is greater than the sum of its
parts.
All systems have elements. These include the goals, as well as the system
inputs, processes, outputs, feedback, and environment.
The structure of systems is defined by its elements and processes. These can
range from simple to complex.
System elements have functional and structural relationships between each
other and are organized in a way to accomplish a specific goal or set of goals.
42


To be part of the system, an element must have a relationship with at least one
element of the system. Any element which has no relationship with any other
element of the system cannot be a part of that system (Von Bertalanffy, 1972;
Hayajneh, 2007, p. 3).
The El system, viewed through the lens of systems theory, is thus a collection of
independent but interrelated organizations that are organized in a meaningful way to
provide early intervention services and supports to children with disabilities or
developmental delays and their families. They are part of the systems resources (which
would include therapists with specialist training); its activities (which are primarily the El
services for children and supports for their families); and its intended goals (that all
children in America are able to reach their full developmental potential) (Funnell, 2000;
Ziviani, Darlington, Feeney, & Head, 2011).
Application to Early Intervention System in Colorado
The El system itself is a component of a larger set of systems, at the nexus of the
health and public health systems, as well as education and human services. Systems
theory highlights these systems do not exist in isolation either; they also function in the
midst of the legislative and judicial systems, the financial and banking systems, and other
systems that comprise the socio-economic-political system of Colorado (Hayajneh,
2007); however, that is beyond the focus of this investigation.
El is an open system, meaning that it interacts with its environment to fulfill the
systems overall goal of providing early intervention services and support to children
with disabilities or developmental delays and their families. Additionally, El is a complex
43


system, meaning that it includes a number of subsystems such as the El programs
network of contracted therapists.
The El system is made up of organized relationships that connect different
elements of the system in specific ways. As detailed in Chapter II, Title V and Part C
authorize and require some of these relationships. Others are formal, but not legislated,
such as the guidance from AAP.
All systems have a goal or set of goals; that is, they must exist for a purpose. The
El system has dual goals of: a) enhancing the development of children with disabilities or
developmental delays from newborn to age three, while also supporting families abilities
to meet their childrens needs; and b) reducing costs to the public education and social
welfare related systems (by reducing the need for special education and minimizing the
likelihood of institutionalization, poverty, and juvenile delinquency and maximizing
independent living and developmental potential (Early Childhood Technical Education
Center [ECTAC], 2008).
Systems theory identifies four interrelated parts of all systems: inputs, processes,
outputs and feedback, leading to the system outcomes (Kast & Rosenzweig, 1972).
These all have attributes that can be measured. Figure III. 1 illustrates the inputs, outputs
and outcomes in the El system in Colorado; along with the factors that influence the
inputs (screening in primary care and/or preschools). Please note that this El system
exists across four separate systems (primary care, public health, human services and
education) as described in Chapter II.
System inputs are the resources used to produce the outputs. In the El system in
Colorado, the inputs include the young children who have developmental delays or
44


disabilities, and their families, along with the therapists with specialized training. All
systems have processes to convert the inputs into outputs. In El, services such as
physical therapy (PT), occupational therapy (OT), and speech therapy are provided.
Outputs are the services or products that result from the system's processes. An output for
the El system in Colorado is that the children receive the support that promotes their
learning and development. Outcomes are the short and long term impacts, as described
previously.
Feedback is information from the system that affects another part of the system.
For example, if staff or contracted therapists do not have adequate skills or expertise
working with children with disabilities or developmental delays, there may be ripple
effects in another part of the system. Further, El is a dynamic system, so it influences
and changes its environment, as it is, in turn, being influenced and changed by its
environment (Hayajneh, 2007).
45


Children with
developmental
delays and their
families.
If:
Developmental
delays are
recognized and
children are
referred for El
services from
primary care
provider,
preschool,
parent, or other.
Figure TTT.1
inputs
Outputs
Outcomes
Long Term
Outcomes
Staff: PT, OT, SP,
social work,
psychology, early
childhood educators,
managers,
administrators.
Staff knowledge,
skills/expertise, time
, and passion, and
' training.
Family insight and
experiences.
Policies, procedures
and guidelines.
Funding for Early
Intervention
Services.
Research on Early
Intervention.

Children receive
supports which
promote their
learning and
development.
Families receive
education and support
regarding their child's
current skills and
ways to promote
further learning and
development
(physical, cognitive,
social and
emotional).
Families receive
skills in advocating
for their child,
supporting the family
unit and promoting
empowerment in the
family.
Community
awareness.

Children have
improved skills and
increased
participation..
Families feel more
confident and
competent in
supporting their
childrens learning
and development.
Families access their
chosen community
services and
activities.
Families experience
being a part of a well
functioning El system
where children and
families needs are
met.
Community members
accept of broader
range of abilities.
Children's
developmental needs
are met, which is
reflected in increased
quality of life, safety,
health and well being.
Families are able to
advocate for
themselves.
Families have higher
expectations and
bigger dreams, and
can participate hi a
full life.
All children are
accepted for who
they are.
Changes to
community attitudes,
legislation, policy
and funding are
evident.
El System Inputs, Outputs and Outcomes (Colorado).
Adapted from Ziviani, et al., 2011.
4^


Finally, the El system in Colorado has functional and structural relationships
among many of the organizations that are organized in a way to accomplish the goals
stated above. It is assumed in systems theory that all elements of the system have
relationships with at least one other element of the system (Von Bertalanffy, 1972). Thus
a focus on the system cannot be limited by the boundaries of each separate organization;
rather it is important to investigate whether there are inefficiencies or opportunities for
improvement in the interorganizational relationships and interdependencies that exist in
such systems.
System of Care Literature
A focus on the system has been emphasized for decades in the system of care
(SOC) literature, which advocates the need for a coordinated network of services and
supports across agencies to meet the multiple and complex needs of children with
disabilities, developmental delays, mental health or other special health care needs
(Stroul, Blau & Friedman, 2010). Developed to help improve the individualized and
collaborative services for children with serious emotional disturbances and their families
(Stroul & Friedman, 1986), the SOC has become a nationally recognized framework for
effective and family-centered service for individuals and families who need services or
resources from multiple human service agencies (Snyder, Lawrence & Dodge, 2012).
The importance of the interorganizational relationships within the system of early
intervention is highlighted when viewed through the SOC framework, as it requires
organizations that provide services and support for children with developmental delays to
communicate, organize, and work together as one coordinated system. In fact, one of the
guiding principles of a system of care is to "ensure that services are integrated at the
47


system level, with linkages between child-serving agencies and programs across
administrative and funding boundaries and mechanisms for system-level management,
coordination, and integrated care management (Stroul, et al., 2010, p. 6). Thus the
relationships between organizations within three silos in Figure II.l (public health,
education and human services and primary care) are critical to ensure that long term
outcomes of the system are achieved. Systems theory and the system of care framework
illustrate that the El system cannot simply be viewed as the delivery of services, as that
view ignores the important interactions between agencies that are so critical for effective
and family-centered services and supports. The systems view also helps to reinforce that
all of the organizations within the El system need to not only deliver quality services,
they also need to work well together to be able to most effectively help and support
children with developmental delays to function better at home, in school, in the
community, and throughout life. It is ultimately about helping children to reach their full
developmental potential.
Overall, there is substantial strength in applying a system approach to better
understand how to improve the health and public health systems. In fact, Mabry and
colleagues (2008), recently stated that the health and well-being of the whole population
may be best conceptualized as a systems problem (p.4). This investigation thus
utilizes systems theory to determine and visualize the parts of the El system in Colorado.
The systems lens helps to illuminate the importance of the relationships among the
different organizations in the context of the system elements: inputs, process, and
outputs. This approach also helps to inform understanding of the important
disconnection between the process of identification and referral of children with
48


developmental disabilities and the legislated processes of the El system (El services and
family supports); and identifies the boundary of the system for this study. Further, the
SOC framework highlights the reasons why such interoganizational relationships are
important. While a valuable heuristic devise, general systems theory has been criticized
for its: vagueness, which has made it difficult operationalize without modifications; and
the assumption that all parts of a system have equal power (Lilienfeld, 1978; Lowe,
1985). The systems approach provides frameworks for conceptualization in this study,
but is not adequate to clearly operationalize and measure key variables in this
investigation, specifically the attributes and characteristics of the stakeholder networks
working to improve the El system. Thus network theories and methodologies are used in
this study to overcome these limitations.
Given this studys focus on system networks, embedding network theories within
the overall theoretical framework of systems theory, particularly in the operationalization
and measurements of the constructs, gives tremendous strength to the theoretical
conceptualization. There are many network theories and measures developed through the
network approach that enable this investigation to measure and link the networks
attributes and characteristics to performance. Such network theories and literature are
described in detail in the following sections.
Network Theories
Scholars from a variety of disciplines have embraced the conceptual notion of
networks to study social and interorganizational relationships. The network approach
builds upon the rational actor assumptions in neo-classical economics, and focuses on the
relationships among and between actors. Informed by the disciplines of mathematics,
49


management, sociology, psychology, and communication, network scholarship has
brought value to these fields, as well as to health, political science, public administration,
management, and public policy, with the development of theory and theoretical
frameworks, modeling tools, analysis techniques and empirical data. Because of its focus
on relationships, this line of research has potential to bring tremendous value to health
and public health systems research, both in theory and practice.
The research on networks makes up a loosely formed, multidisciplinary body of
literature with wide-ranging conceptualizations. Despite the variation in wording,
virtually all definitions of networks contain the common themes of social interaction and
relationships (Provan, Fish & Sydow, 2007). In this investigation, a network is defined
as a group of three or more organizations connected in an effort to achieve or facilitate
achievement of a common goal. The relationships are primarily nonhierarchical among
actors within a network, and may range from loose connections of information flow to
formally integrated services to anywhere in between. It is not only the relationships that
are important in network analyses, but also the absence of relationships, and the
implications of both for achieving the network goals.
Unlike research that focuses on organizations, which is framed by organizational
theory or other traditional guiding framework, there is no one overarching network theory
to guide organizational network scholarship. Instead, researchers from many disciplines
have articulated multiple theories to help explain organizational network structure and
processes. Network research investigates the social structures inherent in health and
public health systems, defined as patterns of connectivity within social systems
(Wellman, 1988, p. 26). Network scholars posit that organizational networks can be more
50


efficient and effective than agencies working in each of their own silos because they can
allow for information and resource exchange, and shared efforts, while reducing
redundancies (Provan & Milward, 1995, 2001; McGuire, 2006; Brass, Galaskiewicz,
Greve & Tsai, 2004; Alter & Hage, 1993; Agranoff & McGuire, 1999, 2001, 2003; Meier
& OToole, 2001). Further, interoganizational networks have been shown to be a
successful tool in building community capacity by leveraging the strengths of diverse
players to solve difficult problems (Agranoff & McGuire, 2003; Provan; Veazie, Staten,
& Teufel-Shone, 2005).
In order to understand the concepts underlying the network approach, it is vital to
have a basic grasp of some of the seminal contributions that build this body of literature,
and to understand different underlying assumptions and levels of analysis. This review
is not an attempt to trace the evolution of all network theories through time, nor is it a
comprehensive review of all network scholarship; rather it focuses on: a) the key facets of
research on individuals social networks that informs research on organizational
networks; and b) organizational network research most relevant to this study and the field
of health and public health systems.
Definition of Network
Network research focuses on the relationships between actors, instead of the
actors themselves. Brass et al., (2004) define a network as a set of nodes and the set of
ties representing some relationship, or lack of relationship, between the nodes (p. 795).
A node is the actor within a network, and can be an individual person, organization or
even a larger group such as a country.
51


In this investigation, organizations are the actors of the stakeholder networks.
The primary caveat is that organizations consist of individuals, thus the social interaction
among organizations ultimately occurs primarily between individuals acting on behalf of
their organizations. Scholars use a variety of terms when referring to the concept of
networks, including partnerships, collaboratives, collations, but nearly all definitions are
centered on the common themes of social interaction, relationships, connectedness,
collaboration, and common goals (Provan et al., 2007).
There are different types of networks, with varying goals and scopes, including,
but not limited to: stakeholder networks, policy networks, service delivery networks,
project based networks, and innovation networks. This study focuses on goal directed
stakeholder networks, comprised of organizations in the community with an interest or
concern with their local system of early intervention. These stakeholders are working
together to overcome the segmentation that exists in the El systems to ensure
standardized developmental screening, referral and follow up for children in their
communities.
Assumptions of Network Theories
Two central assumptions in network theories are important to understand:
embeddedness and social capital.
Assumption of Embeddedness. The study of networks first surfaced in the fields of
anthropology and sociology, as scholars broadened their focus to include the social
context of the actor. The push behind this new line of research was based upon the
assumption that actors do not behave or decide as atoms outside a social context
Granovetter (1985, p. 487). Network theories assume that the rational actors within a
52


network are connected by a set of ties, and that the structure of the ties affects the actors
and their relationships (Coleman, 1988; Moreno, 1934;Nadel, 1957; Granovetter ,1973,
1985, 1992).
Granovetter (1973, 1985) highlighted the first assumption in network research and
began the important trend of distinguishing between types of relationships. He argued
that all economic behavior is embedded in a social context. Granovetters (1973)
strength of weak ties theory, as well as Burts structural holes theory (1992), assume that
patterns of communication and transactions between organizations may depart from an
economic assumption of bounded rationality because actors are embedded in a social
context that influences their decisions and behaviors (Granovetter, 1985). This
assumption is in line with systems theory assertions of interdependence and mutual
influence.
Assumption of Social Capital. Network theories also have an implicit assumption
that relationships are a resource in and of themselves. Coleman (1998) explains that
social capital is
not a single entity, but a variety of different entities having two characteristics in
common: They all consist of some aspect of social structures, and they facilitate
certain actions of individuals who are within the structure. Like other forms of
capital [physical, financial, and human], social capital is productive, making
possible the achievement of certain ends that would not be attainable in its
absence (p. 24).
Unlike economic or human capital, which are attributes of individual actors, social capital
is an attribute of the relationship between two actors (Provan & Lemaire, 2012). Thus,
53


when considering actors within a network, their social capital is the potential access to
information, resources or support inherent in their relationships with the other actors. To
this end, the concept of social capital provides an important frame of reference in its
assertion that relationships are productive resources, worthy of investment. This has
significant application for health and public health systems where a strong return on
investment may be realized from efforts to strategically strengthen interorganizational
relationships for improved system performance.
Theoretical Foundations of the Network Approach
Much of the early theoretical work in network research focused on the social
networks of individuals. Moreno (1934) created the structural approach to social
networks when he developed sociometric models by diagramming relationship networks.
He aimed to identify patterns of interaction, but ended up sparking a whole new way to
examine and frame behavior within social contexts. Not only did a whole new line of
theoretical research emerge, but so too did the current methodological tools of social
network analysis, which map relationships between actors with use of dynamic-network
modeling software packages (Scott, 2000; Hill, 2002). Since network analysis focuses on
relationships between actors, data are generally displayed and analyzed on a matrix of 1
and 0s that indicate the existence or absence of tie.
Nadel (1957) expanded upon Morenos work, arguing the central concept in
sociological theory should be individuals roles. To this end, he explored the
relationships and dynamic between individual actors, their social roles, and the
environment that structures those roles (Nadel, 1957). This new line of research by
Moreno and Nadel was different to earlier trends that focused only on the attributes of the
54


individuals, such as physical or human capital, and the aggregation of those attributes.
Rather, they emphasized the individual relationships and ties among individual actors
within a network, and gave rise to the focus on relationships, instead of on actors, in
theoretical exanimations. While the micro focused social network research seeks to
understand an individuals position in the network for the purpose of maximizing his/her
own self interest; for example, how an actors centrality might affect the power of that
actor; the theories and frameworks have important implications for organizational
network research. Many of the measurements and correlations, referred to as egocentric
in the literature, are applicable and consistency hold true in investigations of relationships
between and among organizations in the literature. With this, scholars crafted a whole
new way to approach investigations in management, organization theory, political
science, policy innovation and diffusion, public management and health policy
implementation.
Organizational Network Research
As the public health system includes all organizations providing public health
services and working toward improved population health, a focus on interorganizational
relationships is critical in health and public health research. This study is interested in
structural and relational patterns that may exist across organizational stakeholder
networks that affect overall network performance. While much of the early network
research focused on individuals as actors within the network, there is a significant body
of literature that focuses on organizations as the actors within network (organizational
networks).
55


Building upon research in intergovernmental relations from political science, and
based upon the assumptions of social network research described earlier, Gage (1984)
was the first public management scholar to investigate inter-government networks. He
sparked a new wave of research investigating the existence and effectiveness of networks
to deliver intergovernmental programs, and helped to launch one of the newer areas of
network research: organizational network analysis (Berry et al, 2004; Agranoff, 1986).
Based upon a review of the literature, many organizational networks share the
following unique features:
They form to achieve a common mission or to deal with complex problems,
which may be difficult to address without collaboration (Berry et al, 2004).
Participation is often voluntary, and is comprised of autonomous actors
(Weiner, Alexander & Zuckerman, 2000).
Network participants often come from diverse organizations, which are
usually geographically distant (Mitchell & Shortell, 2000).
Controlled and regulated by their actors, these networks are usually horizontal
in shape (Alter & Hage, 1993).
There has been a huge growth in research focusing on interorganizational
relations and networks within the field of management in the last couple of decades
(Alter & Hage, 1993; Sydow, 1998). For example, research has shown that embedded ties
of ongoing market relationships affect the cost of capital, client relations, and the
performance of the firms (Uzzi, 1996, 1997, 1999). Other scholars have found that the
repetitive exchanges protect transactions, while reducing transaction costs (Jarillo, 1993).
Without a unifying theory, the body of literature on organizational networks is incredibly
56


diverse. Thus, classifying the organizational network research into categories based upon
level of analysis and focus provides a clarifying framework.
Typology of Organizational Network Research
There are three levels of analysis in an organizational network structure:
1. At the node level of analysis, the actor within a network is the focus. As
discussed, a network actor can be an individual person, organization or a larger group
such as a country; however in organizational network research, the actors are always
organizations. Attributes of organization actors include their sector or field (Borgatti &
Foster, 2003).
2. At the dyadic level of analysis, the focus is the relationship between two
nodes. This is the heart of network research, which focuses on the attributes and
characteristics of the relationships. Many organizational level variables in network
research are based upon dyad data because the relationships belong to the organizations.
3. At the network level of analysis, the focus is on the whole network.
Measurements often include the overall structure of a network, and can include
aggregated normative measurements such as trust. This level is referred to as whole-
network in organizational network literature.
A four by four matrix to classify the different typologies of network research is
illustrated in Table III. 1 Typology of Organizational Network Research. Organizational
network theories actually come from two different, but complementary, perspectives: an
inward perspective and an outward perspective. Borgatti, Jones & Everett (1998)
reviewed the literature and found such a distinction, with scholarship separated by
whether it is focused on looking within the collective or outside the collectivity (p. 29).
57


Wasserman and Galaskiewicz (1994) classify this distinction as a micro-level or macro-
level focus in network research. These distinctions simplify the complex literature, as the
research also varies by the level of analysis: whole network variables or organizational
level variables.
Table IIL1 Typology of Organizational Network Research.
Level of Analysis (IV) Outcome Focus (DV)
Organizational Outcomes (Micro focus) Network Outcomes (Macro focus)
Whole Network Variables Impact of a network on member organizations Impact of network attributes & characteristics on network outcomes.
Organizational Variables Impact of member organizations on other member organizations Impact of organizational/dyadic attributes & characteristics on network outcomes.
Adapted from Provan et al., 2007.
It is important that researchers consider which levels of analysis are most
appropriate for their investigation, and how those fit within the research framing their
investigation. This studys first research question fits into the upper right-hand category
(shaded blue) because it is asking about the impact of whole network variables on
network performance. This study second research question fits into the lower right hand
category (shaded purple) as it is asking about the impact of organizational variables on
network performance.
58


Impact of Whole-Network Attributes and Characteristics on Network Outcomes
This study examines stakeholder networks working to improve the system of
early intervention, first asking: do community based stakeholder networks operating
within the same context exhibit commonalities in whole- network level attributes and
characteristics relative to their performance? This question seeks to better understand
the impact of whole-network characteristics and attributes on performance.
This is an area of considerable interest in health and public health systems
research (Consortium, 2012), but is underdeveloped in the network literature (Provan &
Lemair, 2012). Many organizational network theories draw upon and use many of the
ideas and measures developed by researchers focusing on social networks and ego-centric
measurements; however, the focus is not on the individual actor, but on explaining
properties and characteristics of the network as a whole (NCI, 2007). For example,
instead of examining how organizational centrality might affect the performance of an
actor, this macro-level perspective would focus on how overall network structures and
processes affect network effectiveness. Characteristics and attributes of the network are
used to answer questions such as how overall network performance could be improved.
This perspective assumes that the organizations within a network are working together
toward a common goal.
This area of network research is driven by a quest to understand the structures that
enable networks to be most effective (Provan & Milward, 1991, 1995; OToole, 1997;
Agranoff & McGuire, 1998). Scholars probe a fundamental assumption held by many
policy officials, funders, and service professionals that an integrated network of service
delivery is the most effective approach for providing clients with a continuum of care.
59


This logic can be especially important when serving more vulnerable populationssuch
as individuals with disabilities, the homeless, the elderly, or victims from a natural
disasterwho may have multiple and significant needs or may need help in navigating
complicated and fragmented systems of care.
Network Structure
Researchers have sought to understand to what extent the structural characteristics
of a network, such as density or centrality, may enable networks to be most effective
(Provan & Milward, 1991, 1995; OToole, 1997; Agranoff & McGuire, 1998). For
example, Provan and Milwards (1995) seminal research findings indicate that centrally
coordinated networks may be more effective than decentralized networks in service
delivery networks. The authors found that [njetworks integrated and coordinated
centrally, through a single core agency, are likely to be more effective than dense,
cohesive networks integrated in a decentralized way among the organizational providers
that make up the system (Provan & Milward, 1995, p. 24). Their model of network
effectiveness correlates whole network measures of integration with overall network
effectiveness. Thus, applying Provan and Milwards (1995) model of network
effectiveness to this study context, one would expect to see a fairly clear pattern in this
study that higher performing stakeholder networks would have a higher degree of
centralization and lower performing stakeholder networks would have a lower degree of
centralization. However, subsequent research that focused on the validation of this work
has found contradictory findings (Rosenheck et al. 1998).
Other scholars have sought to understand if there is an ideal level of density for
network effectiveness. Sandstrom and Carlsson (2008) demonstrated that policy
60


implementation is strongest when network density is high. Further, Agranoff and
McGuire (1998) found that the density of networks in local economic development
departments is positively related to achievement of their objectives (adoption of
economic development policy). However, low density/high centrality networks have
been found to be an effective structure for innovation networks, but the relationship
between this whole-network structure and performance has not yet been empirically
tested (Dhanaraj & Parhke, 2006). Provan and Milward (2001) also argue that highly
dense and centralized public service delivery networks work well if the network and
institutional norms support cooperation and collaboration, highlighting the importance of
normative network characteristics as well.
The challenge is that many of these studies varied in their operationalization of
whole-network: some analyzed density (Sandstrom & Carlsson, 2008; Agranoff &
McGuire, 1998), while others analyzed a combination of density and centralization
(Provan & Milwards, 1995; Dhanaraj & Parhke, 2006). Further, the outcome measures
varied significantly as well, leaving this area of organizational network research still
fragmented and under developed. Although there is considerable interest, Provan and
Lemaire (2012) recently stated that there is not yet research available to inform practice
about the right amount of integration in a network (p. 643). Given the lack of a
validated or agreed upon formal theory or model, the studys approach to address the first
research question is exploratory.
Network Trust
While structural measurements are fundamental to network researchers in
understanding the social structures, research has shown that normative measures, such as
61


trust, are also important network characteristics. Network and collaboration research
literature emphasize the importance of trust for effective interorganizational relationships
and networks (Uzzi, 1996; Bryson, Crosby & Stone, 2006; Thomson, Perry & Miller,
2008a). To this end, trust is a common theme in virtually all definitions of
interorganizational networks (Provan et al., 2007, p. 643), with empirical evidence that
overall trust is an integral condition for network performance (Uzzi, 1996; Provan et al.,
2003; Huxham & Vangen, 2005).
Trust is considered a central and ongoing requirement for successful collaboration
because it enables quality work, while reducing transaction costs (Huxham & Vangen,
2005). Putnam (1995) emphasizes the importance of trust to facilitate coordination and
cooperation for mutual benefit (p 67), while Bryson, Crosby and Stone (2006) assert that
trusting relationships are both the lubricant and the glue- they facilitate the work and
they hold collaboration together (p. 48). Trust is also related to network sustainability,
as trusting interorganizational relationships can substitute for legal contracts and formal
organizational agreements on a limited basis, such as between contract periods (Ring &
Van de Ven, 1994; Thomson, Perry & Miller, 2008b). Further, findings in the network
evaluation literature also emphasize that the effectiveness of networks can be influenced
greatly by intra-group dynamics, including trust (Hill, 2002). For example, a network
with low trust may implode before the purpose is achieved. Alternatively, a long
standing network group may have greater trust levels and therefore may be more
effective with limited resources (Hill, 2002).
Although there is little disagreement on the necessity of trust for effective
collaboration, the literature offers many ways to conceptualize it. One recent study in
62


public health networks found that a trusted partner was described as: a) being reliable and
following through; b) sharing the networks goal or mission; and c) being willing to
engage in an open and honest discussion (Varda et al., 2008).
Building upon findings in the literature that overall trust at the network level is
positively correlated with network performance (Uzzi, 1996; Provan et al., 2003), and is
necessary for networks to perform at high levels (Zaheer, McEvily, & Perrone, 1998;
Huxham & Vangen, 2005; Ring & Van de Ven, 1994; Bryson, Crosby & Stone, 2006;
Thomson et al., 2008b), one would expect to see a fairly clear pattern of higher trust in
the higher performing stakeholder networks, and a lower trust in the lower performing
stakeholder networks.
Impact of Organizational Attributes and Characteristics on Network Outcomes
This study also investigates different organizational level attributes and
characteristics, including the connectivity and roles of the organizations; asking in the
second research question: do the relative connectivity and structural positions of member
organizations affect stakeholder network performance? In order to answer this research
question, each organizations dyadic relationships must be analyzed in each stakeholder
network, and then compared across networks to confirm if an association exists between
dyadic relationships and overall network performance.
Organization Positions within Networks
Early network research emphasized the importance of an actors position in a
network. Freeman (1979) examined the role of an individuals position within a network,
and the affect on his or her power. This perspective is also found in structuralist position
theory (Wellman, 1988). Further, Ahuja (2000) found that an actors degree centrality
63


(the number of ties an actor has in a network) increases his information available to him,
while George, Zahra, Wheatley, and Khan (2001) found a positive relationship between
actor degree centrality and the actors capabilities and level of learning.
Burt (1997) also provides an important frame of reference in his assertion that
relationships are not only productive resources, but that well structured networks can be
more effective than large inclusive networks. Further, OToole and Meier (2001) posit a
relationship between network structure and service-delivery performance: they found in
their examination of educational programs that performance improves in districts where
superintendents engage in more network interactions, even if one controls for a variety of
factors that affect this performance (p. 291).
Applying this early network research to organizational networks, an organization
may be in a position of power due to its structural position between two other actors, but
on the other hand, it may have the added responsibility of maintaining the ties between
organizations that are not directly connected in order for the network to function
effectively (NCI, 2007). Identifying organizations that hold a position between other
organizations is one way to identify influential actors in networks (Borgatti et al., 2002).
This may have significant implications for the ability to leverage certain organizations or
structural positions to improve network performance. It would also be important to
consider whether these organizations were spanning structural holes, and thus acting as
brokers in the network as well.
Organization Roles within Networks
Regardless of the overall degree of centralization (or decentralization) of the
network, there are often one or more actors that have greater influence in a network. Burt
64


(1995, 1997, 2000) found that the more areas of the network that an actor has ties with,
the greater the potential information or resources they have access to. Findings in the
literature also indicate that clusters of organizations or fragmentation in a network affect
the flow of communication, resources or knowledge, and may create redundancies or
efficiencies, depending upon the structure and purpose of the network (Burt, 1995). For
example, a fragmented network, that includes unconnected organizations, dyads, and
subgroups, has many structural holes, which has implications for its actors and its
effectiveness. Burts (1995) theory of structural holes posits that there are two
advantages to an actor who bridges the structural holes: a) the actor can control the
information and/or broker the relationships; b) the actor has access to new or non-
redundant information. This theory has been regularly applied to organizational
networks, even though his original research focused on social networks of individuals.
Further, brokerage roles are frequently correlated with influence in an organizational
network (Kilduff & Tsai, 2003).
Kilduff and Tsai (2003) also found that the role of broker in networks is advisable
only for actors who have legitimacy in the social context.... Actors who are considered
outsides, or who are from non-traditional groups, may be punished for attempting to span
across structural holes (p. 58). Given the research setting of El, one would expect that
local health agencies would carry this legitimacy with the members of the stakeholder
networks, enabling the network to use the efficiencies of subgroups, without the loss of
information flow, resources and diffusion. Early childhood education providers and
school districts may also be appropriate brokers, but because they come from education,
rather than health systems, they may be considered outsiders.
65


Connectors within Networks
Krackhardt (1994) also found that the degree of connectedness, which he defined
as the extent to which actors are able to connect to each other through the network, is
an important aspect in network research (in Kilduff & Tsai, 2003, p. 38). Building upon
the importance of connectivity in organizational network research, Varda et al.s (2008)
research into the core connectivity of public health organizations led to a method to
measure the connectivity of organizations that incorporates the quality of their
relationships, as well as the quantity. This is indicated by: a) an organizations centrality;
b) the level of trust that other organizations have for them; and c) how much they value
their relationships with the other organizations (Varda, 2012).
Adding a new consideration to network research, Varda et al. (2008) found that
the most valuable member of the collaborative is considered by public health leaders to
be one that has a credible, well connected presence in the community, can devote
resources (time and money), and contributes the effort to make things happen. (p. E4).
From their findings, the researchers developed three dimensions for assessing the value
of network members: a) power and influence; b) active involvement; and c) contribution
of resources.
Strength of Ties within Networks
Granovetter (1973) posited in his strength of weak ties theory that actors have
strong ties within networks (close relationships) as well as weak ties (acquaintance
relationships); and it is the weak ties that can become a greater resource to the actor. For
example, it is through the weak ties that information is diffused and disseminated and
exchanges are more varied. If one was confined to just a network of strong ties, that
66


actor would be excluded from exchanges in the world around him (Granovetter, 1973).
That said, strong ties also have benefits because they can be trusted sources of influence.
Also differentiating between the types of relationships, Krackhardt (2010) found
that strong ties (as opposed to weak ties) influenced the outcome of a union election
because of the trust inherent in the strong ties; while Hansen (1999) found that strong ties
are more useful to effectively transfer information. Further, Uzzi (1997) found that strong
ties provide greater problem solving capacity. Thus strong ties signal trusted sources of
advice and may be more influential in uncertain or conflicting situation.
However, network research demonstrates that it is unfeasible and unsustainable
for organizations to maintain numerous relationships (Kilduff & Tsai, 2003; Monge &
Contractor, 2003). This would especially be the case for physicians, as their primary
focus is providing individual care. (It is both the focus of their training and the method
through which they are reimbursed for services). This has a significant application in
public health practice, where state and local public health system performance is
assessed, in part, by counting the number of stakeholders in partnerships (NPHPSP,
2007). Through interviews with public health leaders, Varda et al., (2008) found a similar
consensus that it is difficult to maintain interactions between all members of a
collaborative, and that if this was achievable, it probably was not sustainable. (p.E4).
Therefore, they posit that subgroups within a collaborative are efficient... and frequent
interaction is not always necessarily good for connectivity (Varda et al., 2008, p. E4).
Taken together, it is critical that primary care providers are engaged in the network to be
able to bridge the segmentation that exists in El, but simultaneously unfeasible to expect
that they could build and sustain many relationships.
67


Fortunately, network research tells us that being well connected does not mean
that an organization needs to have ties with all members of a network. As with any
research of interorganizational relationships, it is important to consider the context of the
relationships under investigation. In this study, the organizations banded together to
ensure standardized developmental screening, referral and follow up for children in their
communities. These networks were mobilized to bridge a gap between public health,
primary care and education to overcome the segmented system of identification and
delivery for early intervention (El) services. It is vital that primary care physicians are
well connected and engaged in the networks in this study, thus one would expect to see
primary care providers have strong ties with early intervention and public health in the
high performing networks, and weak ties with other organizations in the networks.
Network Analysis in Health and Public Health Systems Research
Network research in public health is generally focused on organizational
networks, with the organizations that make up the networks a blend of government
agencies, other public institutions, private health care providers and non-profit,
community-based organizations. Network research in health and public health is
distinctive from traditional approaches in the field in a number of important ways: first
and foremost, network research focuses on relations and patterns of relations, rather than
on specific attributes of the organizations operating within the health and public health
systems. Second, since network research can be carried out at various levels of analysis,
network analysis can be used to investigate associations between whole network
variables of a health focused network system improvement and/or health outcomes; and
between organizational level variables and system improvement and/or health outcomes,
68


as in this study. Third, network research can integrate both quantitative and qualitative
data, as in this investigation, and thus enable greater richness in the analysis for better
understanding of the affect on health systems or population health outcomes (Kilduff &
Tsai, 2003).
The network approach lends itself very well to health and public health services
and systems research (PHSSR). Given that the public health system includes all
government, private and nonprofit organizations contributing to the core functions of
public health, as described in Chapter II, the network approach offers a promising
approach for capturing the interorganizational relations that form public health systems
(Wholey et al., 2009, p. 184).
Merrill and colleagues (2008) research on local health departments networks
was extended by Wholey et al. (2009), who found that while local health departments
were relatively central, the health departments were usually not the most central
organization in community-based organizational networks. The researchers emphasize
the importance of the networks context, such as community size. Along the same vein,
Mays and colleagues (2006) found that the size of jurisdiction served by a local health
department was a strong predictor of performance, with larger jurisdictions performing
better than their smaller counterparts.
Very little of the network research in health and public health systems research
examines the relationship between network attributes and characteristics and performance
or impact on health outcomes. Suen, Christenson, Cooper and Taylor (1995) showed that
centralized local health departments had better performance measures than those with
decentralized or hybrid structure, but stayed focused on the structure of the organizations,
69


rather than their networks. Hyde and Shortell (2012) found in their systematic review of
public health systems research that the relationships between organizational structure and
health outcomes is complex, calling research that focuses on interorganizational
relationships one of the most notable gaps in the literature. Part of the complexity is that
there is no single organization that can carry out the ten essential public health services
by itself. Archival and survey based studies of public health systems may have an
implicit focus on the local health department as the lead and most central organization,
which may not accurately reflect the system (Wholey et al., 2009). These types of studies
may also be unable to capture the intangible characteristics and attributes of the
relationships within the complex systems.
Depending upon the type of data that is collected, it is possible to examine a range
of questions that shed light on the interorganizational relationships that exist in health and
public health systems, including: the overall level of commitment or influence among
organizations in the network; patterns of involvement; the connection of specific types of
organizations to others; the quality or types interactions and connections; and the
directionality and strength of the relationships; along with the level of trust of
organizations within the network, or the network as a whole. For example, network
research has been used in studies of mental health networks to explore such structural
issues and then compare the findings with those of other networks providing similar
services (NCI, 2007).
The systems level perspective offered by network theories is also important for
health and public health research (Mabry, 2011; Mabry, Olster, Morgan, & Abrams,
2008). Networks are inherent in public health, where public, private and nonprofit
70


organizations work together to assure the conditions for population health. There is a
void of evidence regarding the effectiveness of networks. The grey literature on
partnerships and collaboration in health and public health includes descriptive case
studies of many different initiatives and interventions aimed at improving population
health in some way. However, this collection of descriptive works highlights a
fundamental challenge within this area of research: most partnerships and collaboratives
are locally driven, formed to address a specific barrier, improve access or implement a
specific intervention in their local community. Few studies of networks in health and
public health are generalizable. Significant reasons for this are the unique variations
within the public health systems between, and among, counties and states. Consequently,
this study aims to build upon established systems and network research to investigate
patterns of relationships within the health and public health systems, specifically within
stakeholder networks working to overcome the segmented El system, with the aim to
create new generalizable knowledge.
Network Performance
Although the two bodies of network literature, social networks and organizational
networks, inform this studys operationalization of network performance, there is
additional complexity that must be taken into account when measuring performance in
health and public health. As discussed in Chapter II, early intervention (El) efforts to
offer physical, occupational, and speech therapies to young children with developmental
delays or disabilities have been shown to drastically improve fine and gross motor
development, as well as readiness for school, which then can have positive impacts on the
childs entire life (Jones & Ziglar, 2002; Haskins, 1989). However, for a stakeholder
71


network working within the El system, how can one best measure performance? A count
of the number of children receiving El could be one measure, but that would hardly
capture the complexity of the situation. Who is to say that the network was directly
responsible for each child receiving early intervention? Further, being referred to El may
not guarantee that the child will benefit if other parts of the system do not function
properly. Given the networks goals, how can one measure the impact of El to the lives
of children? If the stakeholder network were responsible for changing just one childs
life, is that not of tremendous value to that family? Does reducing the number of barriers
faced by the families working within the segmented system count as success, even if the
system is still left partially segmented? Would creating awareness about the importance
of standardized screening count as success, even if the increased awareness was not
directly tied to children being referred? All of these questions highlight the tremendous
difficulty in which a statistical count can in no way begin to identify collaboration
success or failure. It is unlikely a single approach to measure the performance of such a
network would be agreed upon in the literature. In fact, some scholars argue that
collaborations should not be identified as successful or not successful unless key
indicators point in that direction over time and in varying contexts (Bingham, 2003).
Review of the network literature shows that scholars are focusing upon the three
structural levels of analysis discussed above- node, dyad and network, but when focusing
on evaluation of performance, some scholars also measure network performance by
looking to community outcomes. For example, Provan and Milward (1995) measured
changes in the incidence of their problem in question and aggregated indicators of client
wellbeing. However, Roussos and Fawcett (2000) lament difficulties with this
72


measurement because visible changes in population-level outcomes take longer than the
lifetime of many [networks] (p. 374).
Consideration of performance literature outside the realm of networks clarifies the
definition of performance measurement. Performance measurement is defined in
business journals as the process of quantifying the efficiency and effectiveness of
action (Neely, Gregory, & Platts, 1995, p. 1229). Thus a performance measure for the
stakeholder networks is a metric or set of metrics that quantify the efficiency and/or
effectiveness of the networks actions (Neely, et al., 1995). The General Accounting
Office (GAO, 2012) defines performance measurement as:
the ongoing monitoring and reporting of program accomplishments, particularly
progress towards pre-established goals. ... Performance measures may address the
type or level of program activities conducted (process), the direct products and
services delivered by a program (outputs), and/or the results of those products and
services (outcomes). A program may be any activity, project, function, or policy
that has an identifiable purpose or set of objectives, (p 3).
Thus applying the two definitions, performance in this study focuses on the most
fundamental question: has the stakeholder network achieved the objectives it was formed
to achieve, addressing both process and outcome measures.
Collaboration researchers often measure the effectiveness as perceived by its
members (Thomson et al., 2008a); while some network scholars have also asked the
network customer and/or client to rate the networks goals achievement (Provan &
Milward, 1995). Responding to the challenges and warning in the literature on
measuring performance, as described in the next chapter on Methodology.
73


Conclusion
Many of the many network attributes and characteristics addressed throughout
this literature review are not apparent to the organizations within a network because they
evolve over time, influenced not only from the decisions of that actor, but also from the
decisions and perceptions of all the other actors within the network. The structures
illuminated by network analysis constitute social realities of which the social actors
themselves may not be aware.... [thus] network research has an emancipatory potential in
that it can inform actors of non-obvious constraints and opportunities inherent in patterns
of social connections (Kilduff & Tsai, 2003, p. 23). Thus answering the research
questions posed in this study requires a structured and robust methodology that draws
upon network analysis instruments, tools and techniques, as detailed in the following
chapter.
74


CHAPTER IV
METHODOLOGY
Research Design
Overview
This chapter provides a detailed description of the design and procedures used in
the study. The methods have a sound epistemological grounding in system and network
theories, and the procedures are consistent with the network approach. Aiming to
understand how patterns of interaction, and other attributes and characteristics of
community-based stakeholder networks affect their performance, this study utilized a
mixed method, most-different cross-case research design.
The research design has two phases. Based on the methodology of the network
approach, Phase 1 focused on the characteristics and attributes of five stakeholder
networks. Whole network and organizational/dyadic variables were examined through
social network analysis (SNA), and analyzed with performance measures based on
stakeholder interviews, self assessments, and local early intervention assessments.
Within case analyses was conducted for each network, and then the findings were further
investigated through cross case analysis to identify whether patterns between independent
variables and performance were observed.
Phase 2 used content analysis to investigate the patterns that emerged from the
cross case analysis, to better understand how network attributes and characteristics affect
network performance. Network findings informed the coding process and the
construction of categories. Specifically, content analysis was used to better understand
75


an emergent distinction in the relationship between key network actors and network
performance.
Each network was treated as a bounded case study, thus allowing for systematic
comparison (Yin, 2003). George and Bennett (2005) argue that there is growing
consensus that the strongest means of drawing inferences from case studies is the use of a
combination of within-case analysis and cross-case comparison within a single case study
or research program (p. 18). As the purpose of the study is to identify patterns in the
attributes and characteristics of networks, relative to their place on a performance
continuum, the cross-case method with a most-different design is a strong choice.
The most-different cross-case research design deliberately seeks to compare cases
that are contrasting or different from each other in order to find similar processes or
outcomes across cases (Przeworski & Teune, 1970). The most different design approach
can indicate the robustness of a relationship that is observed between the independent
variables and the dependent variable. Thus, the approach lends strength to the findings
since such an observed relationship would be consistent across contrasting cases (Faure,
1994). This case-oriented approach thus emphasizes diversity in the selection of cases
(George & Bennett, 2005).
Focusing on the dynamic interactions of public health agencies with public and
private organizations that affect health, the five stakeholder groups are goal-directed
networks, working to ensure standardized developmental screening, referral and follow
up for children in their Colorado communities. The stakeholder networks were mobilized
to overcome the segmented system of early intervention services in their local
76


community. The networks were purposefully selected through key informant meetings to
emphasize diversity in the selection of cases.
Social network analysis can be used to measure variables through confirmed and
unconfirmed ties. Interorganizational relationships are not a one way endeavor, and
analytical methods, such as network analysis capture that complexity.
Identifying the Data Needed
This study investigated both whole network variables and organizational level
variables, as the independent variables, in the context of network performance, the
dependant variable. This study first focused on the impact of whole network variables,
including density, centralization and trust, on network performance, in an exploratory
approach. This studys second question focused on the impact of organizational level
variables, including relative connectivity, brokerage and strength of ties, on network
performance.
Network Performance Data
An understanding of network performance was required for investigation into
both research questions. The stakeholder networks aimed to ensure standardized
developmental screening, referral and follow up for children in their community, with the
ultimate aim to enable Colorados children [to] reach their maximum developmental
potential (ABCD 2012, para. 2). Although it is not possible to measure the networks
exact contribution toward this goal and long term outcome, data on three separate
performance assessments were collected to triangulate performance for each of the
stakeholder networks: an external assessment by ABCD Colorado leadership team;
network members self assessment; and an early intervention assessment. The three
77


separate measurements were used to enhance construct validity and reliability of the
performance measure (Yin, 1994). The rationale and identification of data needed to
answer this studys research questions follows, with the variables detailed in the
Variables and Measurements section later in this chapter.
External Assessment. It is important to understand how the networks were
performing from the perspective of an actor outside of the network, but with an intimate
knowledge of the local El systems and necessary efforts towards systems improvement.
The Assuring Better Child health Development (ABCD) leadership team fit that bill, and
developed the ABCD Model Community Guide: a set of deliverables to help inform
stakeholders efforts and the measure their progress. The ABCD Model Community
guide is a process measure of performance, informed by ABCD Colorados experience
about standardized developmental screening, and the referral and follow-through process
that should occur once children are identified. The guide also takes the factors that need
to be considered when multiple community organizations are performing and advocating
for standardized screenings into account (ABCD, 2011). Given the difficulties in
accomplishing and measuring systems building related efforts, this external assessment
provides an important perspective of network performance. The six key activities and
supporting action steps identified in the guidelines are detailed in the Variables and
Measurements section of this chapter.
Network Self Assessment. Network members own perception of success is
advocated in the literature as another important perceptive of performance. Asking each
member of each stakeholder network to rate how successful the networks focus been for
children in their community captures this important perspective.
78


Early Intervention Assessment. It is also important to capture an objective
assessment in the triangulated performance measure. There is no available data on
screening rates at the community level; however, an indicator that children with
developmental delays or disabilities are being identified early, and the referral and follow
up processes between agencies are working to ensure that children are receiving early
intervention services, was needed. The researcher looked to Early Intervention Colorado
for such an indicator. The degree to which each local community meets the El targets
where the stakeholder networks operates was calculated from the local El reporting data,
and was the third measure used to triangulate network performance.
After identifying the performance data needed, this investigator next focused on
the characteristics and attributes of the networks. This required relational data for each of
the stakeholder networks, meaning that all organizations in each networks needed to
provide information about their relationships with all of the other organizations within the
network.
Whole-Network Data
The studys first research question sought to better understand the impact of
whole-network characteristics and attributes on performance, asking: do community
based stakeholder networks operating within the same context exhibit commonalities in
whole- network level attributes and characteristics relative to their performance? This
is an area of considerable interest in health and public health systems research, but is
underdeveloped in the network literature. Informed by the literature, an exploratory
approach was selected. Network density was identified as one of the whole network
independent variables, with the aim to investigate whether there was a relationship
79


between the overall density of ties in stakeholder networks and network performance.
Further, it was important to understand the extent to which organizations were more or
less connected in the overall network structure. Degree centralization was selected as a
measure to capture the way the network was structured, and how centralized or
decentralized it was, with the aim to investigate whether there was a relationship between
network centralization and network performance (Provan & Lemaire, 2012). These whole
network measures and respective equations are detailed later in the chapter.
Organizational and Dyadic Data
The studys second research question inquired whether any relationships existed
between attributes and characteristics at the organizational level and network
performance. Thus data at the organizational level were identified, collected and
analyzed. Based upon the literature, the organizational level data that capture network
actors structural positions include: degree centrality, relative connectivity, closeness
centrality, betweenness centrality and brokerage relationships (Provan & Lemaire, 2012).
These variables are described in detail in the Variables and Measurements section of this
chapter. Network literature also indicates that a well connected organization is not the
same as the most connected organization. Thus, the type of relationships between
primary care providers and public health and early intervention organizations is
important. For this area of study, it was important that the relational data not simply
indicate an awareness of the organizations or frequency of contact; rather, it needed to
capture the intentional efforts of organizations to enhance the capacity of the primary
care providers to implement and carry out standardized screening and strengthen the
80


follow through process. To this end, the measure of coordinated ties was selected as an
organizational variable and is described in the measurement section.
The Phase 2 content analysis further explored emergent findings from the network
analyses. This mixed method, cross case research design is summarized in Table IV. 1;
with the participants, variables and analysis detailed in the following sections.
Table IV.l Summary of Research Design.
Variables Data Source/ Purpose Analysis
Collection Method Sample
Phase 1
Network Performance (DV) To measure Network
- External Assess - Matrix - ABCD Team and analyze analysis-
- El Assessment - Reports - El Colorado relationships UCINET;
- Self Assessment - Survey - Survey Sample: between whole Within case
All members of network analysis;
Whole Network five purposefully organizational Cross case
Characteristics (IV) selected / attributes and analysis
- Density - Survey stakeholder characteristics
- Deg. Centralization - Survey networks. and
- Trust - Survey performance.
Organizational Characteristics (IV) Centrality Measures - Survey - Survey Sample:
- Rel. Connectivity - Survey All members of
- Brokerage - Survey five purposefully
- Relationship Type - Survey selected stakeholder networks.
Phase 2
Focus of -Organization Mission, vision To gamer in- Content
Organizational websites and value depth data Analysis;
statements of all about Cross Case
organizations emergent Analysis
within all patterns from
networks. analysis.
Data were collected via online surveys from organizations from all five
stakeholder networks detailed in the following sections.
81


Participants and Sites
Focusing on the dynamic interactions of public health agencies with public and
private organizations that affect public health, the population of this study includes five
stakeholder networks working to ensure standardized developmental screening, referral
and follow up for children in their communities. As detailed, these networks were
mobilized to overcome the segmented system of identification and delivery for early
intervention services in Colorado.
Selection of Stakeholder Networks
The five stakeholder networks included in the study were strategically selected
based upon key informant meetings with the Colorado Assuring Better Child health
Development (ABCD) lead team. This investigator employed purposeful sampling of
stakeholder networks guided by theory and specific criteria. Creswell (2007, p. 118)
explains purposeful sampling as the selection of an intentional sample.. .that can best
inform the researcher about the research problem under examination. In this study, each
stakeholder network was selected from specific criteria to ensure it is consistent with
network theories, and to ensure a broad spectrum of performance in which to be able to
compare and contrast across cases.
The first five criteria were based upon Bogason and Zolners (2007) five
characteristics of networks:
1. a relatively stable horizontal articulation of interdependent, but operationally
autonomous actors;
2. who interact through negotiations;
82


Full Text

PAGE 1

i PARTNERS FOR EARLY INTERVENTION AND HEALTHY CHILD DEVELOPMENT: LINKING ATTRIBUTES AND CHARCTERISTICS OF STAKEHOLDER NETWORKS TO PERFORMANCE by Robyn I. Mobbs B.A., University of Colorado Boulder 1999 M.B.A University of Southern Queensland 200 4 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Public Affair s and Administration 2013

PAGE 2

ii This thesis for the Doctor of Philosophy degree by Robyn I Mobbs has been approved for the Public Affairs and Administration Program by Christine Martell, Chair Danielle M. Varda Jessica Sowa Laura Pickler April 4, 2013

PAGE 3

iii Mobbs, Robyn, I. ( Ph.D., Public Affairs ) Partners for Early Intervention and Healthy Child Development: Linking A ttributes and C haracteristics of S takeholder N etworks to P erformance Thesis directed by Associate Professor Christine Martell. ABSTRACT The reliance upon interorganizational relationships in health and public health systems is extensive; however, there is a shortage of evidence that focuses on stakeholder network performance. Responding to this gap and the call for a systems appr oach, the s whether community based stakeholder networks operating within the same context exhibit commonalities in network level attributes and characteristics relative to their performance. The second question focuse s on organizational level variables, asking : do the relative connectivity and structural positions of member organizations affect stakeholder network performance? In a mixed method, cross case research design, the characteristics and attributes of five sta keholder networks wer e analyzed and compared. Whole network, dyad and organizational measures were examined through network analysis, and analyzed with performance measures based on external stakeholder assessments network self assessments and local early intervention assessments. Emergent trends were then analyzed through content analysis. Focusing on the dynamic interactions of public health agencies with public and private organizations that affect health, the five stak eholder groups are goal directed networks working to ensure standardized developmental screening, referral and follow through for children in their Colorado communities. The stakeholder networks were mobilized to overcome the segmented system of early int ervention services.

PAGE 4

iv The study found that there were no relationships in whole network level attributes and characteristics relative to performance Investigation into the second research question revealed a positive relationship between network structur es with a system building organization in a key position relative to network performance; and a negative relationship between network structures with a service provid er organization in a key position relative to network performance There was also a posit ive relationship between the levels of integration of p rimary care providers relative to network performance Results also showed that trust among organizations within a stakeholder network is not a sufficient condition for network performance. The stud y findings provide new evidence about how different characteristics and attributes of stakeholder networks affect performance, adding to a much needed knowledge base to guide the use, formation and management of organizational networks for strong health and public health system s The form and content of this abstract are approved. I recommend its publication. Approved: Christine Martell

PAGE 5

v DEDICATION I dedicate this work to m y husband Simon I cherish and adore you and imagine this wo rld without you. I am grateful for your love and enduring support, and I t your first choice! And to Emily and Jackson: you are both the center of my universe. I love you entirely and completely with all my heart I am grateful for all your love and smiles that have unknowingly seen me through late nights of studying and researching, and kept me moving forward on this journey (Em, I started this PhD the month you were born, and Jack, you entered this world right in the middle of it). You constantly amaze and inspire me, and I hope you know that you are capable of anything (with some hard work and perseverance, of course!). I look f orward to every tomorrow with you. And to my parents, Chris and Kit: you are extraordinary I thank you from the bottom of my hear t for all your love, support wisdom and guidance. I am lucky indeed! And to Ashley: thank you for being the wonderful sister that you are, and for being a great a unt to Em ily and Jack so n so that I could occasionally write in the daylight You are amazing, and inspire me every day with the knowledge that there is al ways a way, it just might mean doing things a little differently! And to Michael : thank you for being an amazing big brother You have pushed and inspired me; encouraged and challenged me thank you! And to Carolyn and Brian, Keith and Lyn Catherine an d Chris and my truly incredible friends : thank you for being you I love and treasure you all.

PAGE 6

ACKNOWLEDGMENTS I would like to thank my amazing committee of incredibly sma rt, furiously hard working, and truly amazing women: Christine Martell Danielle Varda Jessica Sowa and Laura Pickler Christine: thank you for your expertise, guidance and support throughout this entire process. Y ou have been a solid rock for me to lean on and a brilliant advisor proving invaluable feedback and encouragement when I need it most thank you Danielle : thank you for your phenomenal mentorship in public health systems and network research You are a true mentor is every aspect of the word and I am forever grateful. Jessica: your knowledge of the literature and seemingly photographic memory of pertinent studies has been an incredible help to me throughout this endeavor as ha s your assistance and support for teaching. T hank you for e verything. And Laura : you are an outstanding and caring physician and colleague, and you r dedication and abilities to improve the lives of those all around you is truly remarkable. Thank you all so much. I would also like to thank Eileen Bennet and the w onderful and dedicated members of the ABCD team in Colorado I am grateful for your time and support Your efforts are so valuable not only to me during this project, but to all the families and children in the communities that ABCD impacts Thank you. I am also grateful for the funding and ongoing support from the National Coordinating C enter for Public Health Services and Systems Research and the Robert Wood Johnson Foundation. The receipt of the Assuring the Future of Public Health Systems & Services Research: Dissertation Grant Award wa s an honor and provided invaluable financial support. Thank you to the wonderful leadership and staff at both organizations for your confidence in my research, ongoing technical assistance, and

PAGE 7

vii wonderful encourageme nt and inspiration along the way. I feel humbled to be in the company of such great leaders and scholars in PHSSR, and truly appreciate your time and efforts. I am also grateful for the scholarship from Academy Health for the ARM I have such great appre ciation and respect for all that you do in the field and thank you for your support.

PAGE 8

viii TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ .... 1 Background ................................ ................................ ................................ .......... 1 Health Care and Public Health Systems ................................ .......................... 2 Statement of Research Problem ................................ ................................ ............ 4 Interorganizational Relationships ................................ ................................ ... 5 Theoretical Framework ................................ ................................ .................. 6 Research Questions ................................ ................................ .............................. 7 Study Design ................................ ................................ ................................ .. 7 Plan for Dissertation ................................ ................................ ..................... 10 Key Definitions and Terms ................................ ................................ ........... 12 Significance ................................ ................................ ................................ ....... 15 II. RESEARCH CONTEXT: SYSTEM OF EARLY INTERVETION ......................... 17 Background ................................ ................................ ................................ ........ 1 7 Early Intervention Ser vices ................................ ................................ ........... 17 Early Intervention and the Public Health System ................................ .......... 21 National Policy Context: ................................ ................................ ......... 21 State Policy Implementation ................................ ................................ ... 22 Impl ementation in Colorado ................................ ................................ ... 24 Early Intervention and the Education and Human Services Systems ............. 25 National Policy Context ................................ ................................ .......... 25 State Po licy Implementation ................................ ................................ ... 25 Implementation in Colorado ................................ ................................ ... 26

PAGE 9

ix Limitation in Early Intervention System ................................ ................. 27 Health Care System ................................ ................................ ...................... 29 National Conte xt ................................ ................................ .................... 29 Implementation in Colorado ................................ ................................ ... 30 Segmented Early Intervention System ................................ ................................ 30 Identification and Screening: Health care system involvement ..................... 32 Referral to Early Intervention ................................ ................................ ....... 33 Provision of Early Intervention Services ................................ ....................... 34 Implementation Challenges ................................ ................................ .......... 35 Assuring Better Child health Development (ABCD) in Colorado ............ 36 Conclusion ................................ ................................ ................................ ... 38 III. LITERATURE REVIEW ................................ ................................ ....................... 39 Introduction ................................ ................................ ................................ ....... 39 Guided by Theory ................................ ................................ ........................ 39 Research Questions ................................ ................................ ...................... 41 Systems Theory ................................ ................................ ................................ 41 Overarching Theoretical Framework ................................ ............................ 41 Application to Early Intervention System in Colorado ................................ .. 43 System of Care Literature ................................ ................................ ............. 47 Networ k Theories ................................ ................................ ............................... 49 Definition of Network ................................ ................................ .................. 51 Assumptions of Network Theories ................................ ................................ 52 Theoretical Foundations of the Network Approach ................................ ....... 54 Organizational Network Research ................................ ................................ ...... 55 Typology of Organizational Network Research ................................ ............ 57

PAGE 10

x Impact of Whole Network Attributes and Characteristics on Network Outcomes ................................ ................................ ................................ ..... 59 Network Structure ................................ ................................ .................. 60 Network Trust ................................ ................................ ........................ 61 Impact of Organizational Attributes and Characteristics on Network Outcomes ................................ ................................ ................................ ..................... 63 Organization Positions within Networks ................................ ................. 63 Organization Roles within Networks ................................ ...................... 64 Connectors within Networks ................................ ................................ ... 66 Streng th of Ties within Networks ................................ ........................... 66 Network Analysis in Health and Public Health Systems Research ................ 68 Network Performance ................................ ................................ ................... 71 Conclusion ................................ ................................ ................................ ... 74 IV. METHODOLOGY ................................ ................................ ................................ 75 Research Design ................................ ................................ ................................ 75 Overview ................................ ................................ ................................ ..... 75 Identifying the Data Nee ded ................................ ................................ ............... 77 Network Performance Data ................................ ................................ .......... 77 Whole Network Data ................................ ................................ ................... 79 Organizational and Dyadic Data ................................ ................................ ... 80 Participants and Sites ................................ ................................ ......................... 82 Selection of Stakeholder Networks ................................ ............................... 82 Stakeholder Network Members ................................ ................................ .... 85 Census of Ties ................................ ................................ ........................ 88 Human Subjects Review ................................ ................................ ......... 89 Recruitment of Subject Population ................................ .......................... 89

PAGE 11

xi Data Collection ................................ ................................ ............................ 90 Variables and Measures ................................ ................................ ..................... 92 Network Performance Dependent Variable ................................ ................. 92 External Assessment ................................ ................................ ............... 93 Network Self Assessment ................................ ................................ ....... 94 Early Intervention Colorado Assessment ................................ ................ 94 Network Demographics ................................ ................................ ................ 95 Whole Network Independent Variables ................................ ........................ 95 Density ................................ ................................ ................................ ... 95 Degree Centralization ................................ ................................ ............. 96 Network Tr ust ................................ ................................ ........................ 97 Organizational Independent Variables ................................ .......................... 98 Degree Centrality ................................ ................................ ................... 98 Betweenness Centrality ................................ ................................ .......... 99 Brokerage ................................ ................................ ............................. 100 Relative Connectivity ................................ ................................ ........... 100 Relationship Strength ................................ ................................ ........... 102 Data Analysis ................................ ................................ ................................ ... 107 Network Analysis ................................ ................................ ....................... 107 Content Analysis ................................ ................................ ........................ 111 Study Validity and Limitations ................................ ................................ ......... 112 Validity ................................ ................................ ................................ ...... 112 Limitations ................................ ................................ ................................ 114 V. FINDINGS ................................ ................................ ................................ ........... 116 Performance ................................ ................................ ................................ ..... 117

PAGE 12

xii External Assessment ................................ ................................ .................. 117 essment ................................ ................................ ................ 118 Local Early Intervention Assessment ................................ .......................... 119 Demographic Analysis ................................ ................................ ..................... 124 Demographic Variables ................................ ................................ .............. 124 Cross Case Analysis: Demographic Variables ................................ ............ 125 Network Size ................................ ................................ ........................ 126 Network Age ................................ ................................ ........................ 127 Relative County Size (Population) ................................ ........................ 128 Whole Network Cross Case Analysis ................................ ............................... 129 Whole Network Variables ................................ ................................ .......... 129 Density ................................ ................................ ................................ 130 Degree Centralization ................................ ................................ ........... 131 Overall Ne twork Trust ................................ ................................ .......... 132 Network Graphs ................................ ................................ ......................... 133 Cross Case Analysis Organization Level ................................ ......................... 138 Role and Influence of Public Health ................................ ........................... 138 Degree Centrality ................................ ................................ ................. 140 Relative Connectivity ................................ ................................ ........... 140 Closeness and Betweenness Centrality ................................ .................. 142 Brokerage ................................ ................................ ............................. 144 Integration of Primary Care Providers ................................ ........................ 146 Content Analysis ................................ ................................ .............................. 152 Organizational Perspectives: Systems Building vs. Service Provision ......... 152 Conclusion ................................ ................................ ................................ ....... 155

PAGE 13

xiii VI. DISCUSSION ................................ ................................ ................................ ...... 157 Linking Network Attributes and Characteristics to Performance ....................... 157 Summary of Investigation ................................ ................................ .......... 157 Exploring Research Question One ................................ ................................ .... 158 Network Structure ................................ ................................ ...................... 159 Network Trust ................................ ................................ ............................ 161 Exploring Research Question Two ................................ ................................ ... 162 Organization Positions and Roles ................................ ............................... 163 Contributions to the Literature ................................ .............................. 167 Strength of Ties ................................ ................................ .......................... 171 Conclusions fo r Early Intervention, Health and Public Health Systems ............ 174 Conclusion 1: Perspectives of Key Actors Matters ................................ ...... 175 Conclusion 2: Integration of Primary Care can Strengthen Outcomes ......... 175 Conclusion 3: Bigger Is Not Necessarily Better ................................ .......... 177 Future Research Needs ................................ ................................ ..................... 178 REFERENCES ................................ ................................ ................................ ............ 182 APPENDIX ................................ ................................ ................................ ................. 198 1. Early Intervention Process in Colorado ................................ ................................ .... 198 2. Survey Invitation ................................ ................................ ................................ ..... 199 3. Survey Questions ................................ ................................ ................................ ..... 201 4. Stakeholder Network Self Assessment Responses ................................ .................... 207

PAGE 14

xiv LIST OF TABLES T able I.1 U.S. Health Rankings. ................................ ................................ ............................... 1 II.1 Number of Individuals Served and Annual Appropriations for Title V of Social Security Act ................................ ................................ ................................ .................. 24 II.2 Number of Children Served and Annual Appropriations for Part C of IDEA. ......... 26 III.1 Typology of Organizational Network Research. ................................ .................... 58 IV.1 Summary of Research Design. ................................ ................................ .............. 81 IV.2 Assessed Stakeholder Activities. ................................ ................................ ........... 84 IV.3 Conceptual and Operational Definitions Variables and Measures. ..................... 104 V.1 Network Performance. ................................ ................................ ......................... 119 V.2 Summary of Network Demographics. ................................ ................................ .. 125 V.3 Summary of Whole Network Characteristics. ................................ ....................... 129 V.4. Measures of Structural Position and Influence of Network Members. ................... 139 V.5. ECC and PH Relative Con nectivity. ................................ ................................ ..... 145 ................................ ............. 146 V.7 Classifications of Member Organizations. ................................ ............................ 153

PAGE 15

xv LIST OF FIGURES Figure II.1 Systems That Make Up the Early Intervention System at the National, State and Local Levels. ................................ ................................ ................................ ................. 21 II.2 Steps in Early Intervention Process, Cross Listed with System. .............................. 31 III.1 EI System Inputs, Outputs and Outcomes (Colorado). ................................ ........... 46 IV .1 Screenshot of PARTNERTool Survey Login. ................................ ....................... 91 IV.2 Sample Matrix of Stakeholder Network. ................................ ............................. 110 V.1 Comparison of Stakeholder Network Performance. ................................ .............. 122 V.2 Stakeholder Networks along a Performance Continuum. ................................ ...... 124 V.3 Variation in Network Performance by Size. ................................ ......................... 126 V.4 Variation in Network Performance by Age. ................................ .......................... 127 V.5 Variation in Network Performance by Relative County Size. ............................... 128 V.6 Variation in Network Performance by Density. ................................ .................... 130 V.7 Variation in Network Performance by Degree of Centralization. .......................... 131 V.8 Graph of Stakeholder Networ k Alpha. ................................ ................................ 134 V.9 Stakeholder Network Graphs. ................................ ................................ ............... 136 V.10 Network Performance by Relative Connectivity of ECCs. ................................ .. 141 V.11 Network Performance by Relative Connectivity of EIs. ................................ ...... 142 V.12 Network Performance by Brokerage o f EIs. ................................ ....................... 144 V.13 Graph of Stakeholder Network Alpha, Coordinated Activities. ........................... 148 V.14 Stakeholder Networks, Coordinated Activities. ................................ .................. 150 A.1 Self Assessment for Stakeholder Network Alpha. ................................ ................ 207 A.2 Self Assessment for Stakeholder Network Bravo. ................................ ................. 207

PAGE 16

xvi A.3 Self Assessment for Stakeholder Network Charlie. ................................ ............... 207 A.4 Self Assessment for Stakeholder Network Delta. ................................ .................. 208 A.5 Self Assessment for Stakeholder Network Echo. ................................ .................. 208

PAGE 17

xvii LIST OF EQUATIONS Equation IV.1 Density. ................................ ................................ ................................ .......... 96 IV.2 Network Centralization. ................................ ................................ ................. 97 IV.3 Degree Centrality. ................................ ................................ .......................... 99 IV.4 Betweenness Centrality. ................................ ................................ ................. 99

PAGE 18

xviii LIST OF ABBREVIATIONS AB CD Assuring Better Child health Development program BOCES Boards of Cooperative Education CCB Community Centered Board CDC Centers for Disease Control and Prevention CDHS Colorado Department of Human Services CDPHE Colorado Department of Public Health and Environment COMIRB Colorado Multiple Institutional Review Board EI Early Intervention IDEA Individuals with Disabilities Act IOM Institute of Medicine NACHO National Association of County and City Health Officials NIH National Institutes of Health NPHPSP National Public Health P erformance Standards Program Members of the Stakeholder Network s in this Study: PH Local public health department or agency EI Local Community Centered Board, the EI service provider ECC Local Early Childhood Council PC L ocal pediatrician or primary care provider s ECC Local Early Childhood Council EC Local early childhood centers/preschool s BH L ocal b ehavioral h ealth center

PAGE 19

1 CHAPTER I I. INTRODUCTION Background America aims to be "a society in which all people live long, healthy lives" ( United States D epartment of Health and Human Services [HHS] 2012 Message from the Secretary para. 1 is deceptively short and simple; the path towards its achievement is anything but. A healthy society is vitally important to American life: health is a necessity at the individual level for quality of life, at the community level for thriving cities and productive workforces, and at the national level for a growing economy and prosperous citizenry. The problem is that America spends far more than any industrialized nation in the world, yet ranks only mediocre in most important health outcome m easures, as detailed in Table I. 1 below. Ninety seven percent of total health care costs are rela ted to medical care provided on an individual basis ( I nstitute of Medicine [IOM] 2012a ). With costs continuing to increase, there is a consensus among health care and policy leaders that reliance on the primary care model has become fiscally unsustainabl e for the nation ( IOM 2012a ). Further magnifying the challenges of a healthy society is the fact that barriers to health are multifaceted, extensive and interrelated. Table I 1 U.S. Health Rankings. US Rank ing out of total OECD Countries Life Expectancy Infant Mortality Maternal Mortality 26th of 34 30th of 34 (2008 data) 25th of 34 (2007 data) Source : Organization for Economic Co operation and Development [ OECD ] 2009

PAGE 20

2 Health Care and Public Health Systems Ideas to improve the health care system abound, but what has become clear to researchers and policy makers alike is that a systems approach is required. Health care in America is a complex and multifarious set of systems, with layers of subsystems. There are primary care and specialist physicians providing care to individuals, focused for the most part, on treating existing medical symptoms and conditions. A bevy of insurance companies, medical equipment and supply firms, pha rmaceutical companies and billing administrat ors are heavily invested in the provision of care, with their own biases and objectives Hospitals, some for infrastructure, along with government regulators, medical universities and other health training facilities. Research and development, from both the private and public sectors, continue to change the trajectories of the health care business models, clinical services, future research priorities and medical education. Over $2.6 trillion is spent each year on medical care in the health care system described above, and yet that does not represent the whole picture ( IOM, 2012b ). Ensuring a healthy nation clearly requires a system of health care that includes b road access to providers and the provision of quality care, but it also requires efforts to prevent disease and injury; efforts to promote healthy lifestyles; and the assurance of safe conditions and environments in which to live and work (IOM 2012 b ). Thu s, there is the additional layer of the public health system, which is focused on preventing health conditions such as chronic disease, cancer, disability and injury, at the population level (Detels, Beaglehole, Lansang, & Gulliford, 2011) Public health f urther increases the

PAGE 21

3 complexity of health in America bringing a whole new set of actors, funding streams, objectives and level of focus. The public health system in America is anchored by governmental public health organizations, specifically, the 51 sta te public health agencies, 2,794 local public health agencies and 565 American Indian and Alaska Native tribal public health agencies ( National Association of County and City Health Officials [NACHO] 2010 ). These organizations are charged with the task of providing 10 essential public health services (EPHS) deta iled in Key Terms and Definitions in this Chapter However, these governmental organizations do not have the capacity or fun ding to be able to provide the EPHS alone. To fulfill their charge, t hey partner with organizations within all sectors, 2012a p. I 9.) Partners at the national level include federal agencies such as The Health Resources and Services Administration (HRSA), Centers for Disease Control (CDC), and National Institutes of Health (NIH); and national nonprofit organizations such as the Institute of Medicine (IOM) and Robert Wood Johnson Foundation. Partners at the state and com munity level include primary care providers, universities and hospital systems, federally qualified health centers, community based health centers, and mental health centers among others These partners highlight the interwoven complexity and interdepend ency of the health care delivery system with the public health system. Additional state and local partners may include o rganizations outside the realm of primary care, such as business groups and faith based organizations. Community participation is also vital to public health, and thus, local partners also often

PAGE 22

4 include advis ory boards, nonprofit and grass roots organizations as well as families and individuals (IOM 2012a). Traditionally perceived to be two separate systems, there is an increasing call for primary care and public health to integrate because many organizations work within both primary care and public health, sharing the common goal of improving health. The pacity of both sectors to carry out their respective missions and link with other stakeholders to catalyze a collaborative, intersectoral movement toward improv (IOM, 2012 a, p. S 1). It is hoped that better integration will enable both primary care and public health to capitalize on their synergies, and better achieve an overall improvement in heath together. Statement of Research Problem Given the complexity of tasks in assuring conditions for health, and the breadth of organizations t hat are involved, it has been recognized by health leaders and policy makers across the country that change at the system level is required to enable substantial and sustainable improvements in the health of Americans (IOM 2012a). A systems approach, an d an understanding of system level factor s ( and their impact on the delivery of health care and public health services ) is thus critical. Interorganizational relationships make up an important system level factor: partnerships and collaboration are an int egral component of the public health system's organization and structure; an increasingly common method of health and public health service delivery; and an important influence for population health outcomes as an essential public health service unto thems elves (IOM 2011 ).

PAGE 23

5 Interorganizational Relationships Interorganizational relationships are a core feature of the health and public system's organization and structure for the fundamental reason that assuring conditions for population health is beyond the scope and abilities of any one orga nization or governm ent agency (NACHO 2012 ). S takeholder groups and partnerships are the foundation of the loc al public health infrastructure and can include any individual, group, or organization that affects, or can be affected by, actions to improve population health (National Public Health Performance Standards Program [NPHPSP], 2007, p. 24 ) these interorganizational relationships enable public health age ncies to carry out their core functions and essential services. I nterorganizational relationships are clearly important in the health and public health systems where public, private and nonprofit organizations work together to as sure the conditions for he alth; however, there is a significant lack of evidence in this area. The Literature Review in Chapter III chronicles the lack of robust research focused on effective and efficient interorganizational relationships in health and public health systems Fu rther, t he 2012 Public Health S erv ices and S ystems national research agenda created by research scientists and health and public health practitioners and leaders, has identifie d research that focuses on interorganizational relationships in public health as a n important research priority (Consortium 2012). The systems approach enables research that emphasiz es these important interrelations between and among the various actors working in health care and prevention.

PAGE 24

6 Theoretical Framework In the field of s interrelation among them selves and with the environment National Cancer Institute [ NCI ], 2007, p. 1 59 ) Thus, a systems approach requires investigation of the whole, rather than just a n actor or aspect within a system ( Mabry 2011; Mabry Olster, Morgan, & Abrams, 2008; Gillies, 1982). Fundamentally, systems theory emphasizes that a system consists of a set of actors and their relationships. Thus when the US health care and public heal th system s are viewed through the lens of system theory, the relationships between and among the many actors described above are as critical to the system as the actors to which they belong. These relationships can have tremendous impact on the performanc e of the system To this end, there are significant opportunities to maximize and utilize these relationships to leverage the resources, knowledge, information, and services to improve the system and ultimately help to support a nation in which all peop le can live healthier and more productive lives ( United States Department of Health and Human Services [HHS], 2012). The idea that a system is greater than the sum of its parts is at the heart of this approach A systems framework which builds upon s holistic perspective and assumptions of interdependencies is thus a strong research approach to investigate interorganizational relationships embedded in the health and public health systems. Such an approach is further strengthened by th eories of network structure, which aim to foster an understanding of effective relationships among actors and stakeholders to improve collaboration, information and resource flows and exchanges, an d to help reduce

PAGE 25

7 redundancies (Kilduff & Tsai 2003 ). A n ovel methodology i n a systems approach is network analysis, which holds the potential for generating insight and facilitating strategic management of relationships between actors and stakeholders within a system. Many theories of network behavior underpin this type of systems approach and the core network concepts, as detailed in the m ethodology in C hapter IV, have broad acceptance in systems research (NCI 2007 ) The use of network analysi s as a methodology enables exploration of the interorganizational relationships that are so important to the complex health and public health system s and enables investigat ion of unanswered research questions such as those in this study. Research Questions Embracing a systems approach and responding to both the PHSSR research agenda ( S cutchfield, Prez, Monroe & Howard, 2012) and the call to bridge primary care and public health (IOM, 2012a) this study asks: Do community based stakeholder networks operating within the same context exh ibit commonalities in network level attributes and characteristics relative to their performance? The second research question focuses upon organizational level variables, asking : D o the relative connectivity and structural positions of member organization s affect stakeholder network performance ? Study Design In a mixed method, cross case research design, this study investigate s five stakeholder networks working in the context of early childhood intervention (EI), a subsystem that spans the health and public health systems. The EI system at the national, state and local levels is described in Chapter II.

PAGE 26

8 The relational and network measures of the networks will be examined through social network analysis (SNA), and analyzed with performance measures based on external assessments, network self assessments, and early intervention services data Population : Focusing on the dynamic interactions of primary care providers and public health agencies with other community based public and private organizations, the population of this study includes five stakeholder networks working to ensure standardized developmental screening, referral and follow up for children in their local community in Colorado. These networks included stakeholders such as the local public health agency, early intervention service provider, early childhood council and providers, primary care providers, school districts and other organizations i n the community with an interest or concern with the local system of early intervention. These networks were mobilized to overcome the segmented system of identification and delivery for early intervention services. Research Setting: Early intervention is a system of coordinated services that promotes the growth and development of children during the critical early years of brain development. The goal of early i ntervention is to assure that families who have children with disabilities or developmental delays, ages birth to three, receive resources and supports that assist them in maximizing their child's development. Research has shown that s kills and abilities can be fostered and developed during those early years of brain development that may be un attainable if the learning process starts later in life ( Mackrides & Ryherd 2011 ). Utilizing s ystems theory as the general framework and network theories to examine specific measures, this investigation focuses on the interorganizational

PAGE 27

9 relationships that are inherent in each stakeholder network within the EI system that spans the health care delivery system, the public health system, as well as human ser vices and education systems. T his study investigates how different network attributes and characteristics at the network level (referred to hereafter as whole network) and the individual organization/ dyad level ( referred to hereafter as organizational lev el) including structure, connectivity, trust and value affect performance Content analysis is used to further investigate patterns that emerge from the network analysis Theoretical background: While systems theory focuses this investigation on the inte rrelation among and between the multitude of organizations that are a part of the health and public health systems to ensure healthy development; network theor ies provide an important frame of reference in their assertion s that relationships are not only productive resources, but that well structured networks can be more effective than large inclusive networks ( Moynihan, Provan & Lemaire, 2012; Burt 1997). This has a significant application in public health practice, where state and local public health system performance is typically only assessed, in part, by counting the number of stakeholders in partnerships (NPHPSP 2007). Provan, Fish, and Sydow (2007) typology of organizational network research provide a framework f or analysis, while Varda Chandra, Stern (2008) c ore dimensions of connectivi ( 1992, 2010 ) strength of strong ti ) strength of weak ties theory; and (199 5 2000, 2001 ) brokerage research inform the operationalization of network variables to analyze and explore patterns at the whole network and organization level across the five stakeholder networks

PAGE 28

10 Combining the quantitative methodology of social network analysis to measure the social structure, with the exploratory nature of qualitative inquiry to explore emergent patterns, this study aims to provide insight into the dimensions of partnership process and performance. The study findings provide new empirical evidence about how differe nt characteristics and attributes of partnership affect performance, adding to a much needed knowledge base to guide the use, formation and management of interorganizational relationships for effective health and public health system s Plan for Dissertatio n This investigation is presented in six chapters, with the aim to provide insight into the attributes and characteristics of stakeholder networks ( working to improve the system of early intervention ) and their impact on network performance. Chapter II e xplore s the policy context of EI nationally, and describes the segmented system of EI. The chapter describes how the EI system is implemented at the state and local levels in Colorado To this end, it introduces the multiple legislated agencies, along wi th their respective goals, from public health, human services and education, and describes the reasons behind the call for greater integration of primary care providers. The chapter also details the specific type s of organizations that are working together in the five stakeholder networks in this study. Chapter III begins with the literature of systems theory, and offers an application in the context of early intervention. The chapter then navigates the complex body of social and organizational network the ories. Demonstrating a lack of empirical evidence that relate s whole network attributes to network performance, the exploratory approach to the first research question is explained The chapter then synthesizes the network

PAGE 29

11 theories that inform the investigation into the second research question, as well as offer s a discussion on the methodologies and the assumptions contained within. Chapter IV offers a detailed description of all aspects of the research design and methods used in the study inclu ding network analysis to measure th e network attributes and characteristics, and conten t analysis to explore the emergent findings from the cross case analysis The chapter describes the mixed method paradigm utilized in this study, as well as the specific approach and procedures for data collection. It also includes an explanation of the epistemolo gical grounding for the methods and addresses how they are a s trong fit for this line of inquiry The chapter then offers a detailed description of the dimensions of network performance, the dependant variable in this study and the whole network variables and organizational level variables, the independent variable s in this study. The chapter concludes with the limitations of the study, along with the reliability and validity of the approach. Chapter V details the f indings of the network analysis and identifies patterns that exist across the five cases The chapter is organized around the research questions first presenting results from the cross case analysis of the whol e network variables ; and then pr esenting results from the cross case analysis of the attributes and characteristics at the organizational l evel Th e chapter concludes with findings from the content analysis, which was undertaken to better understand a distinction in organizational type that emerged from the cross case analysis. Chapter VI concludes the study with a discussion on the rese arch findings. The first part of this discussion is organized around the research questions, and offers a summary of the findings, interpretation of the data and conclusions drawn from t he

PAGE 30

12 information. A discussion about the contribution of the research to org anizational network theory is offered, as well as the implications for hea lth and public health systems. Four conclusions summarize the contributions to both knowledge and practice. The chapter ends with recommenda tions for future research Key Definitions and Terms System : A system is a collection of independent elements that are interrelated among themselves and within their environment, and organized in a meaningful way to accomplish an overall goal ( Hayajneh 2007). Systems approach : An approach to r esearch or analysis that uses systems theory and methods in an organized framework to investigate systems The National Cancer Institute Initiative on the Study and Implementation of Systems ( NCI, 2007) synthesizes the f our key systems approaches as : 1. Systems organizing to understand and foster the development of participatory, complex, and adaptive collaborative syste ms ; ensure their effective facilitation and management; and encourage productive syst em action and learning. 2. System dynamics to understand and model the complex dynamic interactions involved in the system. 3. System networks to understand and analyze effective collaborative relationships among stakeholders, improve collaboration strateg ies, and help reduce duplication of e ffort. 4. Systems knowledge to develop and manage the knowledge infrastructure required for effective dissemination and evolution of scientifically credible,

PAGE 31

13 evidence based practices, together with an effective strategy to package, deliv er, and maintain this knowledge (p. 2). Systems methods are tools, techniques or procedures to enable investigation of systems. Network analysis, s ystem dynamics modeling, and structured conceptualization are all systems methods (NCI 2 007). Primary Care: Primary care in America is defined by the Inst itute of Medicine (IOM) (1996, p. 1) as the provision of integrated, accessible health care services by clinicians who are accountable for addressing a large majority of personal health care needs, developing a sustained partnership with patients, and practicing in the context of Most prim ary care is delivered through the private sector, but it can also be delivered through government agencies, including the Veterans Health Administration (VHA) hospitals and Health Resources and Services Administration ( HRSA clinics Regardles s of whether it is private or public, primary care physicians often provide a point of first contact for health related issues and aim to offer comprehensive and coordinate d care (Starfield & Horder, 2007). The medical home model of primary care encompas ses the characteristics of care that are considered essential for all children Kogan, & Newacheck, 2011, p. 605). Public Health : ulfilling 140) P ublic health encompasses a diverse group of public and private stakeholders

PAGE 32

14 (including the health care delivery system) working in a variety of ways to contribute to the health of society. Essential Public Health S ervices : The ten Essential Public H ealth S ervices (EPHS) provide a more specific definition of public health and a fr amework for the local public health systems. Each of the state s public health departments and local health agencies are charg ed with carrying out the EPHS. The Center for Disease Control and Prevention (2012) details the EPHS: 1. Monitor health status to identify and solve community health problems. 2. Diagnose and investigate health problems and health hazards in the community. 3. Inform, educate, and empower people about health issues. 4. Mobilize community partnerships and action to identify and solve health pro blems. 5. Develop policies and plans that support individual and community health efforts. 6. Enforce laws and regulations that protect health and ensure safety. 7. Link people to needed personal health services and assure the provision of health care when otherwis e unavailable. 8. Assure competent public and personal health care workforce. 9. Evaluate effectiveness, accessibility, and quality of personal and population based health services. 10. Research for new insights and innovative solutions to health problems ( para 3 ).

PAGE 33

15 Significance Building a health care system that meets the needs of all Americans is an ambitious goal for the government, and requires both comprehensive, well informed policy as well as successful implementation through effective systems Interorgani zational relationships are fundamental in health and public health systems where public, private and nonprofit organizations work together to assure the conditions for population health Yet given the significant reliance on these relationships there is s till much to learn. Uncertainty abounds about when organizational network s are an effective or efficient choice and how to build and foster effective and efficient interorganizational relationships It is important that health practitioners and leaders are armed with the knowledge to prompt strategic and thoughtful discussion to foster stronger inclusion, involvement and commitment where it most makes sense and will have the greatest impact. An understanding of in terorganizational relationships is thus critical in this endeavor and has twice been highlighted as an area of needed research i n the last decade (Scutchfield et al. 2012). The gap in knowledge may exist because traditional research models are challenged by the large and complicated array of actors and variables within health and public health system s and are often not able to take into account that fact that the system s are comprised of interrelated actors and are interrelated themselves Further, the systems are dynamic: they are always changing as feedback from one element of the system changes the course for another. A majority of the research into health care delivery in the twentieth century attempting to understand a problem by first d (NCI, 2007, p.

PAGE 34

16 13; Barabsi, 2002) These traditional approaches have left important research questions unanswered and tremendous gaps in our knowledge of how to improve the health and public health systems in America Given thi s, a systems approach, which guides this study, has gained more attention and traction in the research and practitioner are na s as a way to examine and explore the interrelated elements of the systems and ultimately learn how to better improve and s upport the health of Americans This investigation aims to provide stakeholders the knowledge to begin positioning their stakeholder networks to most effectively achieve their goals and positively influence health outcomes, while maximizing the return on their investment within the networks. Ultimately, stakeholder networks need to be effective in achieving their goals in order to be useful as a tool for policy implementation, and the better we understand the dimensions of interorganizational relationships that affect their success, and how, the more equipped we are to harness and leverage the strengths of these relationships to improve health outcomes in America.

PAGE 35

17 CHAPTER II II. RESEARCH CONTEXT : SYSTEM OF EARLY INTERVETION Background A considerable challenge, with a complex set of systems in p lace to carry out the mission ( HHS, 2012). Inherent in such an endeavor is the important goal of meeting children's hea lth needs, especially for those children who are born with disabilities or those who experience delays in development during their early childhood. The l iterature shows that 12 16 percent of children in America have at least one developmental delay (Mack rides & Ryherd 2011). E arly intervention (EI) system s exist to provide coordinated support s necessary for these children to reach their full potential EI is a system of coordinated services that promotes the child's growth and development and supports families during the critical early years of early brain development. The EI system is especially complex, given that it is legislated at the federal level but implemented at the state and local level s Further, the EI system extends across th e public hea lth system, systems of education and/or human services, depending upon the state, all the while being called to integrate with part of the health care delivery system. S et at the nexus of primary care and public health integration, t his study is focused o n this multifaceted subsystem for childhood development Early Intervention Services The purpose of EI is to ensure that families who have children with disabilities or developmental delays, ages birth to three, receive resources and support that assist them in maximizing their child's development. Early intervention services are focused on the

PAGE 36

18 development that babies typically experience during the first three years of life, including: cognitive development (such as thinking, learning, solvi ng problems); physical skills (reaching, rolling, crawling, and walking); communication; and social development (National Dissemination Center for People with Disabilities [ NICHY], 2012). Early intervention services provide vital support so that children w ith developmental needs can thrive and grow. Services are often provided by physical therapists (PT), occupational therapists (OT), and speech and language pathologists, audiologists, psychologists early childhood educators or nurses These therapists w ork with the children to develop neural pathways, muscles, skills, and knowledge that are not developing in a typical manner, and provide the foundation for language, reasoning, problem solving, behavior, and emotional health. Therapists also provide suppo rt services to the families such as family education and counseling, parent support groups, special instruction, or assistive technology devices and services (NICHY, 2012). There is a substantial body of research that demonstrate s the power and necessity of providing intervention services to children with disabilities and developmental delays at young ages. Research has shown that early intervention services may: Reduce future costs o f services and health care needs (such as special education, therapies or rehabilitation) for both families and government systems; Reduce stress and feelings of frustration for children and families; Help children to reach their maximum potential and bec ome productive, independent individuals (Bright Tots, 2012).

PAGE 37

19 Although there is still much to be learned about the services that work best for different children and families, the combined body of existing research clearly underscored the need to intervene early to enable children with delays and disabilities to reach their full Hebbeler, Spiker, Bailey, Scarborough, Mallik, Simeonsson, et al. 2007, p. 1 1). The benefits of EI services are felt beyond the children and families who receive the m. Research also shows that such services can reduce the prevalence of developmental and behavioral disorders as children get older Further, researchers Hess and Van Landeghem (2005, p. 3) have found that early intervention efforts f or children and their families are not only beneficial and cost effective to the health system, but can also reduce the for more costly interventions and outcomes such as welfare dep I f children do not receive needed services and support they are not only less likely to reach their developmental potential ; there may also be long term consequences for the health, education, child welfare, and justice systems. The importance of early intervention serv ices for infants and toddlers with disabilities is well documented in the literature (Jones & Ziglar 2002; Haskins 1989) ; that is not the investigation here. When children receive EI services, there seems to be little doubt in the research of the positive benefits for the child directly, as well as indirectly to government funded health, education, child welfare, and juvenile justice systems. The problem is that the implementation of the policy is through a segmented system of care. That identification a nd services for EI are divided between public health,

PAGE 38

20 human services education and primary care and resulted in children not always receiving the care they need. Grounded in the health and child development research, three trends at the national level s create d the basis of the early intervention system. Title V of the Social Security Act legislates the public health system to support EI while Part C of the Individuals with Disability Act (IDEA) legislates the systems of education and human s ervices to provide EI services; all the while, medical associatio ns such as American Academy of P ediatrics urge pediatricians and primary care providers to more systematically provide screening and referrals for young children who may need E I services The EI syste m thus spans separate and distinct systems: public health, education human services, and health care (primary care), each with its own unique role to play. Figure II. 1 illustrates the three silos at the national, state and local levels that exist due to the way the services have been legislated. Given this, i nterorganizational relationships are critical in segmented system of services; not only to enable organizations within each system to work together efficiently and effectively but also to brid ge the separate systems These interorganizational relationships may provide the leverage to improve the EI system for the benefit of children and act as a model for the health care and public health systems as a whole.

PAGE 39

21 Figure II 1 Systems That Make Up the Early Inter vention System at the National, State and Local Levels. Early Intervention and the Public Health System National Policy Context : Title V Maternal and Child Health Program of the Social Se curity Act Legislated in 1935 as a part of the Social Security Act, the Title V Maternal and Child Health Program lays a foundation in public health for ensuring the healt h of women, children and youth ( including those with disabilities or special health care needs) in America. Among many things, Title V aims to: Provide and ensure access to preventive and child care services as well as rehabilitative services for certain children; Increase the number of children receiving health assessments and follow up diagnostic and treatment services; Implement family centered, community based, systems of coordinated care for children with special healthcare needs (MCHB 2012a, para. 2 ). Public Health System Social Security Act, Title V State Health Departments (CDPHE in Colorado) Local Health Agencies (55 in Colorado) Education and Human Services Systems Individuals with Disabilities Act, Part C State Departments of Education of Human Services (CDHS in Colorado) Local Human Services or Education Agencies (20 CCBs in Colorado) Health Care System National health professional associations: recomendations and guidance State chapters and medical societies Individual physicians and helath care providers

PAGE 40

22 Title V established the Maternal and C hild Health Bureau (MCHB) at the federal improve the physical and mental health, safety and well being of the maternal and child health (MCH) population which includes all o ad olescents, and their families, i ncluding fathers and children with special health care needs 2012 b para. 2 ) The program plays a critical role in quality oversight, coordination ( between and among progra ms and agencies ), and overall system capacity building (MCHB 2012 b ). State Policy Implementation Maternal and Child Health Services Block Grants (Title V Block grants) are awarded to State Title V programs, whic h are usually located within a m aternal an d child health division of s tate health departments. Each year, t hese programs apply for : State Formula Block Grants, Special Projects of Regional and National Significance (SPRANS), and /or Community Integrated Service Syste ms (CISS) projects (MCHB, 2012). Such implementation allows states and jurisdictions tremendous discretion in how they use of their Title V funds, with the aim to tailor their programs to meet the unique needs and challenges in their states. Each state must identify priorities that add ress the needs of their MCH population, and then guide the use of the Maternal and Child Health Block Grant funds to those ends. To do so requires a partnership arrangement between Federal, State and local entities, and is guided by a state wide needs a ss essment conducted in collaboration with other organizations in the state and local communities every five years. Partnerships are thus a required element of the grant. While discretion

PAGE 41

23 exists e ach state must use its Title V funds to f ocus exclusively on the mater nal and child health population for both infrastructure and service costs Further, the funding requires a match of 75% (meaning that $ 1 of Federal Title V funding must be matched by at least $ .75 of s tate and/or local funding). At least 30 pe rcent of federal Title V funds are earmarked for preventive and primary care services for children ; and at least 30 percent are earmarked for services for children with special health care needs enabling support for early intervention services as necessary The funds may also serve as the payer of last resort for services to children without coverage from another program ( Catalog of Federal Domestic Assistance [CFDA], MCH Block Grants, Number 93.994, 20 12 ) A total of 59 states and jurisdictions (including the District of Columbia and U.S. Territories) receive d Title V Maternal and Child Health Block Grant funding in fiscal year 2011, ranging from $145,927 $ 41,389,219 per state and jurisdicti on, as det ailed in Table II.1 The required match results in almost $1 billion being available annually for maternal an d child health programs at the s tate and local levels. In fiscal year 2011, State Title V programs served over 39 million individuals. Among the i ndividuals served were 2.5 million pregnant women, 4.1 million infants, 27.6 million children, and 1.9 million children with special health care needs ( MCHB, 2012 a ; CFDA MCH Block Grants, Number 93.994, 2012)

PAGE 42

24 Table II 1 Number of Individuals Served and Annual Appropriations for Title V of Social Security Act F ederal Fiscal Year 2011 Number of States and Jurisdictions 59 Number of Individuals Served 39 million Appropriations $546.792 million Source : Catalog of Federal Domestic Assistance 2012 Implementation in Colorado Services receive and distribute the Title V b lock grants to the 55 local public health agencies implementing local The local public health agency staff memb ers are charged with implementing the local action plans in their communities ( Colorado Department of Public Health and Environment [CDPHE], 2012 a ) For example, t he Division of Maternal and Child Health (MCH) at CDPHE is working to improve developmental standardized screening and referral rates for all children, newborn to age 5 and listed it as a 2011 2015 priority ( CDPHE 2012 a ). The Children and Youth Branch (CYB) at CDPHE also receives funding from local sources to help integrate health into local early chil dhood systems building efforts (CDPHE 2012 b ) Title V was a great start, but as more research citing the significant benefits of early intervention was published in the 1970 s and 19 80s, along with a recognized n eed for improved accessibility to education for individuals with disabilities, an EI system of services was legislated in Part C of Individuals with Disabilities Act (IDEA).

PAGE 43

25 Early Intervention and the Education and Human Services Systems National Policy C ontext The overwhelmingly positive findings in the research for early intervention led to legislation mandating EI services in 1986. The Program for Infants and Toddlers with Disabilities (Part C of the Individuals with Disabilities Education Act [IDEA]) legislates state based programs of early intervention services for children aged 0 3 who have developmental delays and disabilities. This federal law was grounded in existing research [that] clearly underscored the need to intervene early to ena ble children et al 2007, p. 1 1). The Education for all Handicapped C hildren Act of 1975 forged the path towards what would eventually become IDEA In this act, federal funded public schools were required to provide equal access to education for children with disabilities. IDEA thus focuses on the educational needs of children with disabilities from birth to age 21, which is why EI services legislated in the Act are provided through e ducation and social services agencies, rather than public health or primary care. The program emphasizes the need for cross agency coordination, again underscoring the need to better understand and improve in terorganizational relationships State Policy I mplementatio n Part C provides federal grants to states to implement early intervention services such as speech therapy, physiotherapy and occupational therapy for eligible children from birth to age three. Again, states have considerable flexibility in use of this funding, which helps the states better meet their own unique needs, but also results in large variations in the proportion of child ren enrolled, ranging from one percent to seven

PAGE 44

26 percent of early childhood populations in each state ( The Early Childhoo d Technical Assistance Center [ ECTAC] 2008 ; Russ, Garro, & Halfon 2010). In fiscal year 2011, nearly 336,895 children under age 3 were served by IDEA Part C programs, as detailed in Table II.2 ( United States Department of Education 2012). Table II 2 Number of Children Served and Annual Appropriations for Part C of IDEA. Federal Fiscal Year 2012 Child Count Year (Fall) 2011 Number of Children Served 336,895 Percent of Population 2.79% Appropriations $442.710 million Source : United States Department of Education 2012 Some states have also been able to expand services to children who are at risk of disabilities or developmental delays, many of whom come from low income families (Russ, Garro, & Halfon 2010). Implementation in Colorad o The Colorado Department of Human Services, Division for Developmental Disabilities administers the Early Intervention Colorado Program and contracts with twenty Community Centered Boards ( CCBs) statewide to provide these early intervention supports and services to children (newborn to age three) and their families (CDHS 2012). The CCBs are responsible for determining eligibility, providing case management and providing or contracting for services and supports (CHDS 2012) The CCBs also serve youth and adults, outside the focus of the EI system.

PAGE 45

27 Limitation in Early Intervention System D espite almost three decades of implementation of PART C of IDEA evidence exists that not all children are receiving the care they need. One of the most significant barriers lies in the challenge in identifying children who have developmental delays during those critical early years of brain development. Kaye, May and Abrams (2006 ) found : many young childr en are not identified with developmental problems until school percent of young children see a child health care clinician in the first three years of life, most of these clini cians are missing opportunities to detect developmental problems, counsel parents of young children about developmental issues, or refer children to ne eded services in the community. (p. 4) These findings are supported by the National Survey of Early Chi ldhood Health, which found that 94 percent of children had parents who were not getting the guidance or education they need as it relates to the sc reening of their child ( Bethell, Reuland, Halfon, & Schor, 2004 ). Dr. Neal Halfon agrees, stating at the Surg eon General's Conference on Children's Mental Health in 2000 children do not receive treatment early in their life un HHS 2005 p. 21 ). Developmental delays are also especially prevalent in young children in low income families, and researchers have found that these children are significantly under detected (Kaye et al ., 2006). In addition to a segmented EI system, America consistently ranks among the lowest in cross c ountry compari sons of child well being (Wise & Blair, 2007). This is

PAGE 46

28 made even more import ant given the body of evidence that demonstrat es the link between adverse early life experiences and development, and poorer health or chronic diseases later in life ( Kuh & Ben Shlomo, 2004 ; Shonkoff & Phillips, 2000 ). Researchers have also demonstrated the powerful links between mental health in childhood and later economic well being (Currie, Stabile, Manivong & Roos 2008). Why are some children not receiving the ca re they need? The systems approach introduced in Chapter I illuminate s the segmentation in the EI system of care. From this perspective, the root of the problem may be that the delivery for early intervention services, which is provided through education and human services agencies, is disconnected with the primary care system, where pediatricians routinely see children birth to three for well child visits and checkups Neither education nor human services would typically have access to children ages birth three, unless they are referred to those systems (Hebbeler et al ., 2007 ). Identification and referral takes place in other systems, specifically the health care syst em, and to some degree, in early childhood education. Early intervention agencies may be well run organizations, but they alone cannot ensure that all children who are in need of services receive them : these agencies are set up to r eact to a referral for services Further, the lead agency that administers the program varies among states, as do the funding mechanism s and the collection of programs that actually provide the services This recognition has led to a call to better integrate the primary care providers in the health care delivery system into the EI system

PAGE 47

29 Health Care System National Contex t The American Academy of Pediatrics (AAP) is a professional membership organization of over 60,000 primary care pediatricians, pediatric medical sub specialists and American Academy of Pediatrics [ AAP ], 2012 Membership, para. 1 ). The American Academy of Family Physicians (AAFP) represe nts over 100,000 family physicians, family medicine residents, and medical students transform health care to achieve optimal health for American Academy of Family Physicians [ AAFP ], 2012 Vision Statement ). These two groups of physi cians make up a considerable part of the health care system in America. The AAP (2012, para. 1 ) health and well such makes recomm endations that often become the bas is of pediatric health care The organization published a report in 2001 highlighting consensus in the field that pediatric clinicians have both the opportunity and expertise to identify children in need of EI services to important role that primary care providers, who see the child on a regular basis and can thus assess development over time, can play in recognizing potential developmental problems Committee on Children with Disabilities, 2001, p p 192 195). The report goes on to assert that pediatricians with both the opportunity and expertise are ideal candidate s for conducting

PAGE 48

30 the ongoi ng surveillance needed to identify developmental problems (AAP Committee on Children with Disabilities 2001). Further, the American Academy of Pediatrics (AAP) also developed the medical home concept, which is becoming the standard for provision of high q uality comprehensive health care (Long, Bauch ner, Sege, Cabral, & Garg, 2012 ). The medical home concept is defined as "model of primary care that is accessible, co ntinuous, comprehensive, family centered, coordinated, compassio nate, and culturally effecti ve" ( Strickland, et al., 2011 p. 605 ). The model requires that children have a consistent health care provider (a physician or nurse) who works with the family to best meet the needs of the child with access to referrals and care coordination as needed (Strickland, et al., 2011). Implementation in Colorado Many of the professional associations such as AAP and AAFP have active state chapters that support individual pediatricians and family medicine physicians practicing in Colorado whether they are based in private partnerships or solo practices throughout the state or work for federally funded VA clinics or HRSA safety net clinics The Colorado Medical Society and other state based organizations also play a key role in the primary care system in C olorado, as do physician assistant and nurse practitioner organizations. Segmented Early Intervention System As mentioned above, t here are separate steps in the EI process, each taking part in a different system First and foremost is identificat ion and r eferral, represented as red in Figure II.2 which often takes place within a primary care setting when initiated by a

PAGE 49

31 physici an or referred to a physician, or within public health through a safety net clinic or home visitation program. There is then the eligibility process that is facilitated by the CCBs on behalf of human services. This may also include screening if not conducted by a physician beforehand If deemed eligible the CCB provides the EI services represented in purple in the figure below or contracts therapists to provide the services On the child system to the education system, represented by the blue arrow below. Figure II 2 Steps in Early Intervention Process C ross L isted with System. Mackrides and Ryherd (2011, p. 544) found that while 12 16 percent of children in the United States have a developmental delay, as many as one half of affected children will not be i delays are not detected until they are five, opportunities for developmental gains through early intervention are lost. Further, without well defined inter agency and inter system processes there are opportunities for a child to be accidentally dropped between the syst ems. Finally, a system perspective highlights an extremely important facet of the EI system, that no one agency or system can serve the child alone: supporting a child with a Identification and Referral Health Care System/ Public Health System/ Early Education Elgibility Provision of EI Services Human Services System Education Services Education System

PAGE 50

32 di sability or developmental delay to reach their full health and developmental potential requires successful action in all parts of the systems. Identification and Screening: Health care system involvement While some families may know that early interventio n services will be essential in helping the ir child grow and develop from the time of birth (if the child is born with a specific disability that is diagnosed at the hospital for example ), other children may have a relatively routine birth, and then devel op more slowly or quite differently than other children. These children require someone to identify their developmental delay(s) a nd refer them for EI services tool wil l more effectively identify children who may be at risk for, or have, a developmental delay than physicians who do not use su Shevell, Majnemer & Oskoui, 2005, p. 4). Several studies also indicate that using a developmental screen ing tool with decisions based only on clinical judgment al., 2005, p. 8 ). However, there are indications that pediatricians and family medicine physicians do not regularly use standar dized tools health visits; inadequate compensation; lack of training in the use of specific tools; and 9; HHS, 2000 ) Although most physicians report using informal developmental checklists as part of their overall strategy of care, a literature review by National Academy of State Health Policy found indications that a low percentage of children in need of support for

PAGE 51

33 their healthy development are identified, even by physicians ( Sices, 2007; Rydz, et al., 2005). The literature on clini cal recommendations for screening in primary care is inconsistent and often insufficient to direct the primary care physician. In addition, multiple barriers exist, including time limited appointments and lack of resources which often prevent physicians f rom performing initial screening and completing additional evaluation and referrals (Mackrides & Ryherd, 2011 ; Sices, 2007 ). If a child starts at elementary school with a n undiagnosed disability or developmental delay, the child stays in the education system to receive special education services ; however, a critical period of brain development has already passed, and the child may not be able to make the therapeutic or clinical advances that might have been possible through early interven tion (NICHY 2012). Referral to E arly Intervention When viewing early intervention through a systems approach, it becomes clear that improving the screening and referral of children with developmental delays will help to overcome some of the segmentation in the system. However, even if screening improves through actions of the pediatricians and primary care providers, it will do little good for the child if there is no coordination between the health care system and human services and education systems. Further to this point, research has shown that providing families and clinicians access to resources for assessment and treatment are also critical (Kaye, 2006) (or continue) usin g a screening tool unless they [are] confident that the children they

PAGE 52

34 identified as potentially needing further care would receive appropriate care (Kaye, 2006 p 26 ). Provision of Early Intervention Services As mentioned, the Colorado Department of Human Services, Division for Developmental Disabilities administers the Early Intervention Colorado Program in Colorado. It provide s early intervention supports and services through contracts with twenty Community Centered Boards (CCBs) across the state, which in turn, provide the services to infants, toddlers, and their families within their communities. The EI system in Colorado is visually depict ed in Appendix 1. It is clear that the EI system in Colorado crosses many systems: p ublic health has worked for y ears to ensur e care through maternal and child health programs and initiatives as funding from Title V programs at the national level is provided to CDPHE, who in turn, provide s it to local health agencies throughout the state. With the creation of IDEA, the services for EI are formally legislated and funded, but administered within multiple systems: in Colorado, they are housed within human services and education systems. Yet, the important time for EI services is b irth to age 3 a time before children start with the education system. Human services may not have involvement life at all, unless referred. And thus the EI system is also dependent upon the health care system, with a significant reliance on primary care providers to screen and identify children with developmental delays. Further, there is a new acknowledgement that early childhood directors and teachers can also help to indentify children with devel opmental delays.

PAGE 53

35 e prosperity in the world, and the substantial health care expenditures for seniors (via Medicare), the government makes only a relatively modest investment in early childhood development (Bennett, 2008a & Bennett, 2008b). Through the health and p ublic he alth systems that exist and with coordinati on with human and educational systems, Am erica must be able to leverage funding and resources to be as effective as possible for our young children As discussed, in terorganizational relationships are a critical element in this endeavor Implementation Challenges In summary, systematic reviews of the research on early intervention have shown that EI services can enhance physical, emotional, and psychological growth for children. Such services have also demonstrated positive short term and long term effects for the child, family and government systems (Barnett, 1995; Jones & Zigle r, 2002 ). The liter ature also illustrates the highly segmented structure in the delivery of early intervention services, which are provided through education and human service sectors, while recent consensus emphasizes the parallel role of screening and identification as par t of primary medical care ( King & Glascoe 2003; Nelson ). This popularity of the growing research on early brain development, which places child develo pment in a biomedi cal context ( Berry, Kutz, Langner & Budetti 2008, p. 481 ). The recognition that the health care system may represent a more logical first point of contact for infants and young children ( than education and h uman service s ) has gained considerable strength, and further emphasize s

PAGE 54

36 the need t o str engthen and leverage the ties between primary care and public health and between primary care/public health and education and human services systems Recognizing thes e significant issues, and to stimulate innovative models for integrating child development and medical care, the Commonwealth Fund a private foundation dedicated to improving the health care system, sponsors the Assuring Better Child Health and Development (ABCD) initiative. The initiative assists states in improving the delivery of e arly child development services by strengthening primary health care services and systems that support the healthy development of children from newborn to age three (Berry et al., 2008). The ABCD program focuses on building low ., 200 8, p. 32). Sponsored by t he Commonwealth Fund and administered by the National Academy for State Health Policy (NASHP), the program has been operating since 2000 in various ing National Academy for State Health Policy [NASHP], 2012, para. 1). Ass uring Better Child health Development (ABCD) in Colorado r maximum D, 2012 p ara 1 ). This guides its the use of standardized developmental screening tools in health care settings across Colorado to facilitate early identifi p ar a 2). program then received funding to expand outreach to the entire state two years later, and

PAGE 55

37 a state team was developed to guide the efforts. The team includes pediatric primary health care providers, parents, and leaders from state organizations and health care systems. ABCD Colorado has worked with primary health care providers in over 40 of 2012) since 2005. ABCD Colorado aims to support health care physicians by promoting early identification of young children with potential developmental delays and referral to appropriate community services and resources. It assist s primary care practices in implementing st andardized developmental screening in an efficient and practical way; and aim s to help a physician pr actice s build and strengthen relationships with early intervention resources and services in their community (ABCD, 2012b). It also provide s training to ot her professionals who work with children during the critical ages from newborn to age three, including child care and early education providers. This might include training on developmental milestones standardized developmental screening tools, and appro priate referral processes (ABCD, 2012c). ABCD Colorado support ed community driven stakeholder networks to form in Five of these networks are being investigated in this study. The stakeholder networks are working to ensure standardized developmental screening, referral and follow up for children in their own local community in Colorado. While each group stakeholder networks in this study include the following organizations and agencies each described in greater detail in Chapter IV: Local public health department or agency (PH)

PAGE 56

38 Local Community Centered Board, the EI service provider contracted by the Colorado department of Humans services (EI) Local school district or Board of Cooperative Education (ED) Local pediatrician or primary care provider (PC) Local Early Childhood Council (ECC) Local early childhood centers/preschool (EC) Local behavioral health center (BH) Conclusion In an effort to ensur e that all children are able to reach their maximum development, the stakeholder networks operating within the EI system provide an excellent set of cases to investigate, as the local implementation must be collaborative: many organizations must work toget her with a number of government agencies, private health care practices, and community based non profits to overcome the limitations of a segmented system The ABCD program has a grant funded state team to guide the program, and with its leadership, it su pport s community driven, autonomous stakeholder ensure that every child in Colorado has the opportunity to reach their maximum developmental potential (ABCD, 201 2 ) This investigation will examine and ana lyze multiple variables of process, structure and comprehensive analysis of five stakeholder networks working to improve the EI system in Colorado. The segmentatio n and limitations of the EI system lends to the importance of being able to leverage interorganizational relationships to improve system capacity and

PAGE 57

39 CHAPTER III III. LITERATURE REVIEW Introduction The legislative context and policy implementation of early intervention, as described in the preceding chapter, leads to a system where no single organization can ensure that children receive necessary early intervention services. An effective and efficient system of EI services in Colorado requires primary care providers and early childhood educators to identify and refer children who may benefit and be eligible for EI services. It also requires that the system incorporate feedback and follow through on referrals, a focus of public health. Further, i t requires human services and education to coordinate an effective transition every time a child turns four and move s from one system to the next for continue d services. An effective EI system thus require a number of organizations and agencies with varied goals, resources, knowledge, and information to work together ; highlighting the need for effective interoganizational relationships. Guided by Theory This study examines the questi ons of networks working to improve their local EI system s and how the attributes and characteristics of the interorgan izational relationships within stakeholder network s impact performance. Grounded in a systems approach, this study accepts the fundamental assumption of systems theory that a system is greater th an the sum of its parts, and requires investigation of the whole, rather than specific aspects of a system, issue or problem ( V on Bertalanffy, 1972 ) Thus this investigation focuses on the interrelation of

PAGE 58

40 organizations within the system of EI, which exi sts within the larger context of the health and public health system s More specifically, this approach utilizes the concepts in network theories and the applications of network analysis to ultimately examine how different attributes and characteristics of interorganizational relat ionships affect the performance of stakeholder network s While the theoretical lens of a systems approach offers great value in its holistic perspective of illuminating how and why a system is more than the sum of its parts, syst ems theory requires additional theoretical support in a mixed, method cross case research design, such as this one (NCI, 2007). Network theories, which focus on relationships between and among actors, are powerful in their mathematical examination of relational structure and ties. Network research offers a collection of theories and frameworks that correlate the processes through which relationships can exert influence on outcomes at the individual, organizational, or network level ( Gulati, Nohria & Zaheer et al., 2000 ). Although missing an overarching theory, network research fundamentally prove s that social structure matters ( Ahuja, Soda & Zaheer 2012). N etwork research in the interorganizational context (hereafter referred to as the network a pproach), is increasingly popular in health and public health research, and provides strength in undertaking and explaining not only the interaction, connection and structure of organizations partnering in health and public health, but also their affect on performance and health outcomes (Wholey Gregg & Moscovice 2009 ; Mabry, 2011 ). Combined, these frameworks offer a powerful theoretical and methodological base for this study B y embedding the network approach within the overarching system theory, one can leverage the strengths of each framework, while minimizing their

PAGE 59

41 respective weaknesses. To this end, the literature review will reintroduce the research questions, follow with a discussion of systems theory and then explore the network approach. Re search Questions With an emphasis on interorganizational relationships in health and public health s and performance this study aims to answer the following research questions: d o community based stakeholder networks operating within the same context exhibit commonalities in whole network level attributes and characteristics relative to their performance? The second research question focuses upon organizational level attributes and characteristics asking: do the relative connectivity and structural positions of member organizations affect stakeholder network performance ? As described, this study examines stakeholder networks working in the context of early interven tion, a system that spans health and public health systems as well as human services and education systems Systems Theory Overarching Theoretical Framework The f oundation for systems theory has a long and storied history, traceable back more importance than material nature, known in modern times as the simple principle that the whole is greater than the sum of its parts ( NCI 2007 ) General sy stems theory was developed by Ludwig von Bertalanffy (1972 ), a biologist who crafted the theory to make sense of system characteristics such as wholeness, differentiation, and order (Gillies 1982). These roots of systems theory have

PAGE 60

42 informed new theories including chao s theory and nonlinear dynamics, relational mathematics, game theory, sys tem dynamics; and network theories Systems theory frames the world as a complex system composed of subsystems that interact with each other. Each subsystem has cle arly defined b o undaries and coherent dynamics ( V on Bertalanffy, 1972 ). The theory can be used to understand the EI system, including its structures, processes and outcomes and their interactions with the health care and public health systems. Systems theory is thus used in this investigation as a n overarching theoretical framework to indentify the elements of the EI system; the relati onships between these elements; and th s N etwork theories are then used to identify and operationalize the attributes and characteristics of the stakeholder networks that include the actors within the system working to improve the system System Characteristics. According to systems theory, most systems have the following common characteristics: The elements of a system comprise a whole that is greater than the sum of its parts. All systems have elements. These include the goal s as well as the system input s process es output s feedback, and environment. The structure of systems is defined by its elements and processes. These can range from simple to complex. System elements have functional and structural relationships between each other and are organized in a way to accomplish a specific go al or set of goals.

PAGE 61

43 To be part of the system an element must have a relationship with at least one element of the system. Any element which has no relationship with any other element of the system cannot be a part of that system ( Von Bertalanffy, 1972 ; Ha yajneh, 2007, p. 3 ). The EI system viewed through the lens of systems theory, is thus a collection of independent but interrelated organizations that are organized in a meaningful way to provide early intervention services and supports to children with disabilities or developmental delays and their families They are part of the urces ( which would include therapists with specialist training ) ; its activities ( which are primarily the EI services for children and supports for their families); a nd its intended goals (that all children in America are able to reach their full developmental potential) (Funnell, 2000; Ziviani, Darlington, Feeney, & Head, 2011 ). Application to Early Intervention System in Colorado The EI system itself is a component of a larger set of systems, at the nexus of the health and public health systems, as well as education and human services Systems theory highlight s these systems d o not exist in isolation either; they also function in the midst of the legislative and ju dicial systems, the financial and banking systems, and other systems that comprise the socio economic political system of Colorado (Hayajneh, 2007); h owever, that is beyond the focus of this investigation. EI is an open system, meaning that it interacts with its environment to fulfill the overall goal of providing early intervention services and support to children with disabilities or developmental delays and their families Additionally, EI is a complex

PAGE 62

44 system, meaning that it include s a number o f subsystems such as the EI network of contracted therapists The EI system is made up of organized relations hips that connect different elements of the system in specific ways. As detailed in Chapter II, Title V and Part C authorize and require some of these relationships Others are formal, but not legislated such as the guidance fro m AAP All systems have a goal or set of goals; that is, they must exist for a purpose. The EI system has dual goal s of : a) enhancing the development of children with disabilities or developmental de lays from newborn to age three, while also supporting famil abilit ies needs ; and b) reducing costs to the public education and social welfare related systems ( by reducing the nee d for special education and minimizing the likelihood of institutionalization, poverty, and juvenile delinquency and maximizing independent living and developmental potential (Early Childhood Technical Education Center [ ECTAC ], 20 08 ) Systems theory identi fies four inte rrelated parts of all systems: inputs, processes, outputs and feedback, leading to the system outc omes (Kast & Rosenzweig, 1972 ). These all have attributes that can be measured. Figure III.1 illustrates the inputs, outputs and outcomes in t he EI system in Colorado ; along with the factors that influence the input s (screen ing in primary care and/or preschools ) Please n ote that this EI system exists across four separate systems (primary care, public health human services and education) as described in Chapter II. System inputs are the resources used to produce the outputs. In the EI system in Colorado the inputs include the young children who have developmental delays or

PAGE 63

45 disabilities, and their families along with the therapists with specialized training All systems have processes to convert the inputs into outputs. In EI, services such as physical therapy ( PT ) occupational therapy ( OT ) and speech therapy are provided Output s are the service s or product s that result from the syst em's processes An output for the EI system in Colorado is that t he children receive the support that promote s their learning and development Outcomes are the short and long term impacts as described previously Feedback is information from the system that affects another part of the system. For example, if staff or contracted therapists do not have adequate skills or e xpertise working with children with disabilities or developmental delays there may be ripple e ffects in another part of the system Further, EI is a dynamic system, so it influenc es and chang es its environment as it is in turn being influenced and changed by its environment (Hayajneh, 2007).

PAGE 64

46 Adapted from Ziviani, et al., 2011. Figure III 1 EI System Inputs, Outputs and Outcomes (Colorado) I nput s Staff: PT, OT, SP, social work, psychology, early childhood educators, managers, administrators. Staff knowledge, skills/expertise, time and passion, and training. Family insight and experiences. Policies, procedures and guidelines. Funding for Early Intervention Services. Research on Early Intervention. Outputs Children receive supports which promote their learning and development. Families receive education and support regarding their child's current skills and ways to promote further learning and development (physical, cognitive, social and emotional). Families receive skills in advocating for their child, supporting the family unit and promoting empowerment in the family. Community awareness. Outcomes Children have improved skills and increased participation.. Families feel more confident and competent in supporting their and development. Families access their chosen community services and activities. Families experience being a part of a well functioning EI system where children and families needs are met. Community members accept of broader range of abilities. Long Term Outcomes Children's developmental needs are met, which is reflected in increased quality of life, safety, health and well being. Families are able to advocate for themselves. Families have higher expectations and bigger dreams, and can participate in a full life. All children are accepted for who they are. Changes to community attitudes, legislation, policy and funding are evident. 55 Children with developmental del ays and their families If: Developmental delays are recognized and children are referred for EI services from primary care provider, preschool, parent, or other.

PAGE 65

47 Finally, the EI system in Colorado has functional and structural relationships among many of the organizations that are organized in a w ay to accomplish the goals stated above. It is assumed in systems theory that all elements of the system have relationships with at least o ne other element of the system ( Von Bertalanffy, 1972 ). Thus a focus on the system cannot be limited by the boundaries of each separate organization; rather it is important to investigate whether there are inefficiencies or opportunities for improvement in the interorganizational relationships and interdependencies th at exist in such systems System of Care Literature A focus on the system has been emphasized for decades in the system of c are (SOC) literature which advocates the need for a coordinated network of services and supports across agencies to meet the multip le and complex needs of children with disabilities, developmental delays, mental health or other special health care needs (Stroul, Blau & Friedman, 2010). Developed to help improve the individualized and collaborative services for children with serious e motional disturbances and their families (Stroul & Friedman, 1986), the SOC has become a nationally recognized framework for effective and family centered service for individuals and families who need services or resources from multiple human service agenc ies (Snyder, Lawrence & Dodge, 2012). The importance of the interorganizational relationships within the system of early intervention is highlighted when viewed through the SOC framework, as it requires organizations that provide services and support for children with developmental delays to communicate, organize, and work together as one coordinated system. In fact, o ne of the guiding principles of a system of care is to "ensure that services are integrated at the

PAGE 66

48 system level, with linkages between chil d serving agencies and programs across administrative and funding boundaries and mechanisms for system level management, the relationships between organizations within three silos in Figure II.1 (public health, education and human services and primary care) are critical to ensure that long term outcomes of the system are achieved. Systems theory and the system of care framework illustrate that t he EI system cannot simply be v iewed as the delivery of services, as that view ignores the important interactions between agencies that are so critical for effective and family centered services and supports. The systems view also helps to reinforce that all of the organizations within the EI system need to not only deliver quality services, they also need to work well together to be able to most effectively help and support children with developmental delays to function better at home, in school, in the community, and throughout life. It is ultimately about helping children to reach their full developmental potential. Overall, t here is substantial strength in applying a system approach to better understand how to improve the health and public health systems. In fact, Mabry and colleagues ( 2008 ) recently stated being of the whole population (p.4). This investigation thus utilizes systems theory to determine and visualize the parts of the EI system in Colorado The systems lens helps to illuminate the importance of the relationships among the different organizations in the context of the system elements : inputs, process, and outputs. This approa ch also helps to inform understanding of the important disconnection between the process of identification and referral of children with

PAGE 67

49 developmental disabilities and the legislated processes of the EI system (EI services and family supports) ; and identif ies the boundary of the system for this study Further, the SOC framework highlights the reasons why such interoganizational relationships are important. While a valuable heuristic devise, general systems theory has been criticized for its : vague ness wh ich has made it difficult operationalize without modifications ; and the assumption that all parts of a system have equal power (Lilienfeld, 1978; Lowe, 1985) The systems approach provides framework s for conceptualization in this study, but is not adequat e to clearly operationalize and measure key variables in this investigation specifically the attributes and characteristics of the stakeholder networks working to improve the EI system. Thus network theories and methodologies are used in this study to overcome these limitations embedding network theories within the overall theoretical framework of systems theory, particularly in the operationalization and measurements of the construct s gives tremendous str ength to the theoretical conceptualization There are many network theories and measures developed through the network approach that enable this investigation to measure and attributes and characteristics to performa nce Such network t heories and literature are described in detail in the following section s Network Theories Scholars from a variety of disciplines have embraced the conceptual notion of networks to study social and interorganizational relationships. The network approach builds upon the rational actor assumptions in neo classical economics, and focuses on the relationships among and between actors. Informed by the disciplines of mathematics,

PAGE 68

50 managem ent, sociology, psychology and communication, network scholarship has brought value to the se fields, as well as to health, political science, public administration, management and public policy with the development of theory and theoretical frameworks, modeling tools, analysis techniques and empirical data. Because of its focus on relationships, t his line of research has potential t o bring tremendous value to health and public health systems research, both in theory and practice. The research on network s makes up a loosely formed, multidisciplinary body of literature with wide ranging conceptualizations. Despite the variation in wording, virtually all definitions of networks contain the common themes of social interaction and relationship s (Provan, Fish & Sydow, 2007) In this investigation, a network is defined as a group of three or more organizations connected in an effort to achieve or facilitate achievement of a common goal. The relationships are primarily nonhierarchical among actors within a netw ork, and may range from loose connection s of information flow to formally integrated services to anywhere in between. It is not only the relationships that are important in network analyses, but also the absence of relationships, and the implications of b oth for achieving the network goals. Unlike research that focuses on organizations, which is framed by organizational theory or othe r traditional guiding framework there is no one overarching network theory to gu ide organizational network scholarship. Ins tead, res earchers from many disciplines have articulated multiple theories to help explain organizational network structure and processes. Network research investigates the social structures inherent in health and public health (Wellman, 1988, p 26). Network scholars posit that organizational network s can be more

PAGE 69

51 efficient and effective than agencies working in each of their own silos because they can allow for information and resource exchange, and shared efforts while reducing redundancies (Provan & Milward 1995, 2001; McGuire 2006; Brass Galaskiewicz, Greve & Tsai 2004; Alter & Hage 1993 ; Agranoff & McGuire 1999, 2001, 2003 ; Meier & 2001). Further, interoganizational networks have been shown to be a successful tool in building community capacity by leveraging the strengths of diverse players to solve difficult problems (Agranoff & McGuire 2003; Provan ; Veazie, Staten, & Teufel Shone 2005). In order to understand th e concepts underlying the network approach, it is vital to have a basic grasp of some of the seminal contributions that build this body of literature, and to understand different underlying assumptions and levels of analysis. This review is not an attemp t to trace the evolution of all network theories through time, nor is it a comprehensive review of all network scholarship; rather it focuse s on: a) the key facets of social networks that informs research on organizational networks ; and b) organizational network research most relevant to this study and the field of health and public health systems Definition of Network Network research focuses on the relationships between actors, instead of the actors themselves. Brass et al., (2 004 (p. 795). A node is the actor within a network, and can be an individual person, organization or even a larger group su ch as a country

PAGE 70

52 In this investigation, organizations are the actors of the stakeholder networks. The primary caveat is that organizations consist of individuals, thus the social interaction among organizations ultimately occurs primarily between indivi duals acting on behalf of their organizations. Scholars use a variety of terms when referring to the concept of network s including partnerships, collaboratives, collations, but nearly all definitions are centered on the common themes of social interactio n, relationships connectedness, collaboration and common goals (Provan et al. 2007). There are different types of networks with varying goals and scopes, including, but not limited to: stakeholder networks, policy networks, service delivery networks, project based networks, and innovation networks. This study focuses on goal directed stakeholder networks comprised of organizations in the community with an interest or concern with their local system of early intervention. These stakeholders are work ing together to overcome the segmentation that exists in the EI systems to ensure standardized developmental screening, referral and follow up for children in their communit ies Assumptions of Network Theories T wo central assumptions in network theories are important to understand: embeddedness and social capital. Assumption of E mbeddedness The study of networks first surfaced in the fields of anthropology and sociology, as scholars broadened their focus to include the social context of the actor. The push behind this new line of research was based upon the Granovetter (1985, p 487). Network theories assume that the rational actors within a

PAGE 71

53 network are connected by a se t of ties, and that the structure of the ties affects the actors and their relationships (Coleman 1988; Moreno 1934; Nadel 1957; Granovetter 1973, 1985, 1992 ) Granovetter (1973 1985) highlighted the first assumption in network research and began the important trend of distinguishing between types of relationships. He argued that all economic behavior is embedded in a social context 73 ) strength of weak ties theory, as well as Burt 1992) assume that patter ns of communication and transactions between organizations may depart from an economic assumption of bounded rationality because actors are embedded in a social context that influences their decisions and behaviors (Gra novetter, 1985). This assumption is i influence Assumption of S ocial C apital Network theories also have an implicit assumption that relationships are a resource in and of themselves Coleman (1998 ) explains that social capital is not a single entity, but a variety of different entities having two characteristics in common: They all consist of some aspect of social structures, and they facilitate certain actions of individuals who are within the stru cture. Like other forms of capital [physical, financial, and human], social capital is productive, making possible the achievement of certain ends that would n ot be attainable in its absence (p 24). Unlike economic or human capital, which are attributes of individual actor s social capital is an attribute of the relationship between two actor s (Provan & Lemaire, 2012). Thus

PAGE 72

54 when considering actor s within a network, their social capital is the potential access to information, resource s or support inherent in their relation ships with the other actors To this end, the concept of social capital provides an important frame of reference in its assertion that relationships are productive reso urces, worthy of investment. This has significant application for hea lth and public health systems where a strong return on investment may be realized from effort s to strategically strengthen interorganiza tional relationships for improved system performance. Theoretical Foundations of the Network Approach Much of the early theoretical work in network research focused on the social networks of individuals. Moreno (1934) created the structural approach to social networks when he developed sociometric models by diagramming relationship networks. He aimed to identify patterns of interaction, but ended up sparking a whole new wa y to examine and frame behavior within social context s. Not only did a whole new line of theoretical research emerge, but so too did the current methodological tools of soc ial network analysis, which map relationships between actors with use of dynamic network modeling software packages (Scott, 2000; Hill, 2002). Since network analysis focuses on relationships between actors, data are generally displayed and analyzed on a matrix of 1 and 0s that indicate the existence or absence of tie. To this end, he explored the relationship s and dynamic between individual actors, their social rol es, and the environment that structure s those roles (Nadel, 1957). This new line of research by Moreno and Nadel was different to earlier trends that focused only on the attributes of the

PAGE 73

55 individuals such as physical or human capital and the aggregation of those attributes. Rather, they emphasized the individual relationships and ties among individual actors within a network and gave rise to the focus on relationships, instead of on actors, in theoretical exanimations. While the micro focused social network research seeks to actor; the theories and frameworks have important i mplications for organizational network research. Many of the measurements and correlations referred to as ego centric in the literature, are applicable and consistency hold true in investigations of relationships between and among organizations in the li terature. With this, scholars crafted a whole new way to approach investigations in management, organization theory, political science, policy innovation and diffusion, public management and health policy implement ation Organizational Network Research As the pub lic health system includes all organizations providing public health services and working toward improved population health a focus on interorganizational relationships is critical in health and public health research. This study is interested in structural and relational patterns that may exist across organizational stakeholder networks that affect overall network performance. While much of the early network research focused on individuals as actors within the network, there is a significant body of literature that focuses on organizations as the actors within network (organizational networks).

PAGE 74

56 Building upon research in intergovernmental relations from political science, and based upon the assumptions of social networ k research described earlier, Gage (1984) was the first public management scholar to investigate inter government networks. He sparked a new wave of research investigating the existence and effectiveness of networks to deliver intergovernmental programs, and helped to launch one of the newer areas of network research: organizational networ k analysis (Berry et al, 2004; Agranoff, 1986 ). Based upon a review of the literature, many organizational network s share the following unique features: They form to a chieve a common mission or to deal with complex problems, which may be difficult to address without collaboration (Berry et al, 2004) Participation is often voluntary, and is comprised of autonomous actors (Weiner Alexander & Zuckerman 2000). Network participants often come from diverse organizations, which are usually geographically distant (Mitchell & Shortell, 2000). Controlled and regulated by their actors, these networks are usually horizontal in shape (Alter & Hage, 1993). There has been a huge growth in research focusing on interorganizational relations and networks within the field of management in the last couple of decades (Alter & Hage, 1993 ; Sydow, 1998). For example, r esearch has shown that embedded ties of ongoing market r elationships affect the cost of capital, client relations, and the performance of the firms (Uzzi, 1996, 1997, 1999). Other scholars have found that the repetitive exchanges protect transactions, while reducing transaction costs ( Jarillo, 19 93 ). Without a unifying theory, the body of literature on organizational networks is incredibly

PAGE 75

57 diverse. Thus, classifying the organizational network research into categories based upon level of analysis and focus provides a clarifying framework. Typology of Organiza tional Network Research There are three levels of analysis in an organizational network structure : 1. At the node level of analysis, the actor within a network is the focus. As discussed, a network actor can be an individual person, organization or a larger group such as a coun try; however in organizational network research, the actors are always organizations. Attributes of organization actors include their sector or fiel d (Borgatti & Foster, 2003) 2. At the d yadic level of analysis, the focus i s the relationship between two nodes. This is the heart of network research which focuses on the attributes and characteristics of the relationships. Many organizational level variables in network research are based upon dyad data because the relationsh ips belong to the organizations. 3. At the network level of analysis the focus is on the whole network Measurements often include the overall structure of a network, and can include aggregated normative measurements such as trust. This level is referr A four by four matrix to classify the different typologies of network research is illustrated in Table III.1 Typology of Organizational Network Research. Organizational n etwork theories actual ly come from two different, but complementary, perspectives: an inward perspective and an outward perspective. Borgatti Jones & Everett (1998) reviewed the literature and found such a distinction, with scholarship separated by

PAGE 76

58 Wasserman and Galaskiewicz (1994) classify this distinction as a micro level or macro level focus in network research. T hese distinctions simpl if y the complex literature, as the research also varies by the level of analysis: whole network variables or organizational level variables Table III 1 Typology of Organizational Network Research. Level of Analysis (IV) Outcome Focus (DV) Organizational Outcomes (Micro focus) Network Outcomes (Macro focus) Whole Network Variables Impact of a network on member organizations Impact of network attributes & characteristics on network outcomes. Organization al Variables Impact of member organizations on other member organizations Impact of organizational/dyadic attributes & characteristics on network outcomes. Adapted from Provan et al., 2007 It is important that researchers consider which levels of analysis are most appropriate for their investigation, and how those fit within the research framing their investigation. fits into the upper right hand category (shaded blue) because it is asking about the impact of whole network variables on second research question fits into the lower right hand category (shaded purple) as it is asking about the impact of organizational variabl es on network performance

PAGE 77

59 Impact of Whole Network Attributes and Characteristics on Network Outcomes This study examines stakeholder networks working to improve the system of early intervention, first asking: do community based stakeholder networks ope rating within the same context exhibit commonalities in whole network level attributes and characteristics relative to their performance? This question seeks to better und erstand the impact of whole network characteristics and attributes on performance. This is an area of considerable interest in health and public health systems research (Consortium, 2012), but is underdeveloped in the network literature (Provan & Lemair, 2012). Many organizational network theories draw upon and use many of the ideas and measures developed by researchers focusing on social networks and ego centric measurements; however, the focus is not on the individual actor, but on explaining properties and characteristics of the network as a whole (NCI, 2007). For example, instead of examining how organizational centrality might affect the performance of an actor, this macro level perspective would focus on how overall network structures and processes affect network effectiveness. Characteristics and attributes of the network are u sed to answer questions such as how overall network performance could be improved. This perspective assumes that the organizations within a network are working together toward a common goal This area of network research is driven by a quest to understand the structures that enable networks to be most effective (Provan & Agranoff & McGuire, 1998). S cholars probe a fundamental assumption held by many policy officials, funders, and service professionals that an integrated network of service delivery is the most effective approach for providing clients with a continuum of care.

PAGE 78

60 This logic can be especia lly important when serving more vulnerable populations such as individuals with disabilities, the homeless, the elderly, or victims from a natural disaster who may have multiple and significant needs or may need help in navigating complicated and fragment ed systems of care. Network Structure Researchers have sought to understand to what extent the structural characteristics of a network, such as density or centrality, may enable networks to be most effective (Provan & Agranoff & McGuire, 1998). For coordinated networks may be more effective than decentralized networks in service delivery networks. The authors found that integrated and coordinated centrally, through a single core agency, are likely to be more effective than dense, cohesive networks integrated in a decentralized way among the organizational pr oviders that make up the system Provan & Milward, 1995, p. 24). Their model of network effectiveness correlates whole network measures of integration with overall network effectiveness to this study context, one would expect to see a fairly c lear pattern in this study that higher performing stakeholder networks would have a higher degree of centralization and lower performing stakeholder networks would have a lower degree of centralization. However, subsequent research that focused on the val idation of this work has found contradictory findings ( Rosenheck et al. 1998 ). Other scholars have sought to understand if there is an ideal level of density for network effectiveness. Sandstrom and Carlsson (2008) demonstrated that policy

PAGE 79

61 implementatio n is strongest when network density is high. Further, Agranoff and McGuire (1998) found that the density of networks in local economic development departments is positively related to achievement of their objectives (adoption of economic development policy ). However, low density/high centrality networks have been found to be an effective structure for innovation networks, but the relationship between this whole network structure and performance has not yet be en empirically tested (Dhanaraj & Parhke, 2006). Provan and Milward (2001) also argue that highly dense and centralized public service delivery networks work well if the network and institutional norms support cooperation and collaboration, highlighting the importance of normative network characteristi cs as well. The challenge is that many of these studies varied in their operationalization of whole network: some analyzed density (Sandstrom & Carlsson, 2008; Agranoff & McGuire, 1998), while others analyzed a combination of density and centralization ( Provan & ; Dhanaraj & Parhke, 2006 ). Further, the outcome measure s varied significantly as well leaving this area of organizational network research still fragmented and under developed. A lthough there is considerable interest, Provan and Lemaire (2012) recently stated that there is not yet (p. 643 ) Given the lack of a validat ed or agreed upon e ss the first research question is exploratory Network Trust While structural measurements are fundamental to network researchers in understanding the social structures, research has shown that normative measure s such as

PAGE 80

62 trust, are also important network characteristic s Network and collaboration research literature emphasize the importance of trust for effective interorganizational relationships and networks (Uzzi, 1996; Bryson, Crosby & Stone, 2006; Thomson, Perry & Miller, 2008 a ) To this end, trust is overall trust is an integral condition for network performance (Uzzi, 1996; Prova n et al., 2003 ; Huxham & Vangen, 200 5). T rust is considered a central and ongoing requirement for successful collaboration because it enables quality work, while reducing transaction costs (Huxham & Vangen, 2005 and they facilitate the work and Trust is also related to ne twork sustainability, as trusting interorganizational relationships can substitute for legal contracts and formal organizational agreements on a limited basis, such as between contract periods (Ring & Van de Ven, 1994; Thomson, Perry & Miller, 2008 b ). Fur ther, f indings in the network evaluation literature also emphasize that the effectiveness of network s can be influenced greatly by intra group dynamics including trust (Hill, 2002). For example, a network with low trust may implode before the purpose is achieved Alternatively, a long standing network group may have greater trust levels and therefore may be more effective with limited resources (Hill, 2002). Although there is little disagreement on the necessity of trust for effective collaboration, th e literature offers many ways to conceptualize it. One recent study in

PAGE 81

63 public health networks found that a trusted partner was described as: a) being r eliable and following through engage in an open and honest discussion (Varda et al., 2008). Building upon findings in the literature that overall trust at the network level is positively correlated with network performance ( Uzzi, 1996; Provan et al., 2003 ), and is necessary for networks to pe rform at high levels ( Zaheer, McEvily, & Perrone 1998; Huxham & Vangen, 2005; Ring & Van de Ven, 1994; Bryson, Crosby & Stone, 2006; Thomson et al. 2008 b ), one would expect to see a fairly clear pattern of higher trust in the higher performing stakeholder networks, and a lower trust in the lower performing stakeholder networks. Impact of Organizational Attributes and Characteristics on Network Outcomes T his study also investigates different organizational level attributes and characteristics including the connectivity and roles of the organizations; asking in the second research question: do the relative connectivity and structural positions of member or ganizations affect stakeholder network performance ? In order to answer this research question, each network and then compared across networks to confirm if an association exists bet ween dyadic relationships and overall network performance. Organization Positions within Network s Early network research emphasized the importance of position in a network Freeman (1979) and the affect on his or her power. This perspective is also found in structuralist position

PAGE 82

64 (the number o f ties an actor has in a network) increases his information available to him, while George Zahra, Wheatley, and Khan (2001) found a positive relationship between Burt (1997) also provides an important frame of reference in his assertion that relationships are not only productive resources, but that well structured networks can be a relationship between network structure and service delivery performance: they found in superintendents engage in more network interactions, even if one controls for a variety of Applying this early network research to organizational networks, an organization may be in a position of power due to its structural position between two other actors, but on the other hand, it m ay have the added responsibility of maintaining the ties between organizations that are not directly connected in order for the network to function effectively (NCI, 2007). Identifying organizations that hold a position between other organizations is one way to identify influential actors in networks (Borgatti et al., 2002). This may have significant implications for the ability to leverage certain organizations or structural positions to improve net work performance. It would also be important to conside r whether these organizations were spanning structural holes, and thus acting as brokers in the network as well. Organization Roles within Network s Regardless of the overall degree of centralization (or decentralization) of the network there are often one or more actors that have greater influence in a network. Burt

PAGE 83

65 (199 5 1997, 2000) found that the more areas of the network that an actor has ties with, the greater the potential information or resources they have access to. Findings in the literature als o indicate that clusters of organizations or fragmentation in a network affect the flow of communication, resources or knowledge, and may create redundancies or efficiencies, depending upon the structure and purpose of the network (Burt, 199 5 ). For exampl e, a fragmented network, that includes unconnected organizations, dyads, and subgroups, has many structural holes, which has implications for its actors and its 5 ) theory of structural holes posits that there are two advantages t o an actor who bridges the structural holes: a) the actor can control the information and/or broker the relationships; b) the actor has access to new or non redundant information. This theory has been regularly applied to organizational networks, even tho ugh his original research focused on social networks of individuals. Further, brokerage roles are frequently correlated with influence in an organizational network (Kilduff & Tsai, 2003). Kilduff and Tsai (2003) also found that the role of broker in networks is advisable . Actors who are considered outsides, or who are from non traditional groups, may be punished for attempting to span tting of EI, one would expect that local health agencies would carry this legitimacy with the members of the stakeholder networks, enabling the network to use the efficiencies of subgroups, without the loss of information flow, resources and diffusion. Ea rly childhood education providers and school districts may also be appropriate brokers, but because they come from education, rather than health systems, they may be considered outside r s.

PAGE 84

66 Connectors within Network s Krackhardt (1994) also found that the de gree of connectedness, which he defined an important aspect in network research ( in Kilduff & Tsai, 2003, p. 38). Building upon the importance of connectivity in org anizational network research research into the core connectivity of public health organizations led to a method to measure the connectivity of organizations that incorporates the quality of their relationships, as well as the quantit y This b) the level of trust that other organizations have for them ; and c) how much they value their relationships with the other organizations (Varda, 2012) Adding a new consideration to network research, Varda et al. (2008) found that the most valuable member of the collaborative is considered by public health leaders to resourc From their findings, the researchers developed three dimensions for assessing the value of network members: a) power and influence; b) active involvement; and c) contribution of resources. Strength of Ties within Network s Granovetter (1973) posited in his strength of weak ties theory that actors have relationships); and it is the weak ties that can become a greater resource to the actor. For example, it is through the weak ties that information is diffused and disseminated and exchanges are more varied. If one was confined to just a network of strong ties, that

PAGE 85

67 actor would be excluded fro m exchanges in the world around him (Granovetter, 1973). That said, strong ties also have benefits because they can be trusted sources of influence. Also differentiating between the types of relationships, Krackhardt (20 10 ) found that strong ties (as opp osed to weak ties) influenced the outcome of a union election because of the trust inherent in the strong ties; while Hansen (1999) found that strong ties are more useful to effectively transfer information. Further, Uzzi (1997) found that strong ties prov ide greater problem solving capacity. Thus strong ties signal trusted sources of advice and may be more influential in uncertain or conflicting situation. However, network research demonstrates that it is unfeasible and unsustainable for organizations to m aintain numerous relationships (Kilduff & Tsai, 2003; Mong e & Contractor, 2003 ). This would especially be the case for physicians, as their primary focus is providing individual care. (It is both the focus of their training and the method through which th ey are reimbursed for services). This has a significant application in public health practice, where state and local public health system performance is assessed, in part, by counting the number of stakeholders in partnerships (NPHPSP, 2007). Through inter views with public health lea ders, Varda et al., (2008 ) found a similar collaborative, and that if this was achievable, it probably was not sustainable Therefore th Taken together, it is critical that primary care providers are engaged in the network to b e able to bridge the segmentation that exists in EI, but simultaneously unfeasible to expect that they could build and sustain many relationships.

PAGE 86

68 Fortunately, network research tells us that being well connected does not mean that an organization needs to have ties with all members of a network. As with any research of interorganizational relationships, it is important to consider the context of the relationships under investigation. In this study, the organizations banded together to ensure standardiz ed developmental screening, referral and follow up for children in their communities. These networks were mobilized to bridge a gap between public health, primary care and education to overcome the segmented system of identification and delivery for early intervention (EI) services. It is vital that primary care physicians are well connected and engaged in the networks in this study, thus one would expect to see primary care providers have strong ties with early intervention and public health in the high p erforming networks and weak ties with other organizations in the networks Network Analysis in Health and Public Health Systems Research Network research in public health is generally focused on organizational network s, with the organizations that make up the networks a blend of government agencies, other public institutions, private health care providers and non profit, community based organizations. Network research in health and public health is distinctive from tr aditional approaches in the field in a number of important ways: first and foremost, network research focuses on relations and patterns of relations rather than on specific attributes of the organizations operating within the health and public h ealth syst ems. Second, since network research can be carried out at various levels of analysis, network analysis can be used to investigate associations between whole network variables of a health focused network system improvement and/or health outcomes ; and betwe en organizational level variables and system improvement and/or health outcomes

PAGE 87

69 as in this study Third, network research can integrate both quantitative and qualitative data, as in this investigation, and thus enabl e greater richness in the analysis for better understanding of the affect on health systems or population health outcomes (Kilduff & Tsai, 2003). The network approach lends itself very well to health and public health services and systems research (PHSSR) Given that the public health system includes all government, private and nonprofit organizations contributing to the core functions of approach for capturing the interorganizational relations that form public (Wholey et al., 2009 p. 18 4). networks was extended by Wholey et al. (2009), who found that while local health departments were relatively central, the health departme nts were usually not the most central organization in community based organizational network s. The researchers emphasize Mays and colleagues (2006) found that the size o f jurisdiction served by a local health department was a strong predictor of performance, with larger jurisdictions performing better than their smaller counterparts. Very little of the network research in health and public health systems research exam ines the relationship between network attributes and characteristics and performance or impact on health outcomes. Suen Christenson, Cooper and Taylor (1995) showed that centralized local health departments had better performance measures than those with decentralized or hybrid structure, but stayed focused on the structure of the organizations,

PAGE 88

70 rather than their networks. Hyde and Shortell (2012) found in their systematic review of public health systems research that the relationships between organizatio nal structure and health outcomes is complex, calling research that focuses on interorganizational relationships one of the most notable gaps in the literature. Part of the complexity is that there is no single organiza tion that can carry out the ten esse ntial public h ealth s ervices by itself. Archival and survey based studies of public health systems may have an implicit focus on the local health department as the lead and most central organization, which may not accurately reflect the system (Wholey et al., 2009). These types of studies may also be unable to capture the intangible characteristics and attributes of the relationships within the complex systems. Depending upon the type of data that is collected, it is possible to examine a range of q uestions that shed light on the interorganizational relationships that exist in health and public health systems, including : the overall level of commitment or influence among organizations in the network; patterns of involvement; the connection of specifi c types of organizations to others; the quality or types interactions and connections; and the directionality and strength of the relationships; along with the level of trust of organizations within the network, or the network as a whole. For example, net work research has been used in studies of mental health networks to explore such structural issues and then compare the findings with those of other networks providing similar services (NCI, 2007). The systems level perspective offer ed by network theories is also important for health and public health research (Mabry, 2011; Mabry, Olster, Morgan, & Abrams, 2008) Networks are inherent in public health, where public, private and nonprofit

PAGE 89

71 organizations work together to assure the conditions for population h ealth. There is a void of evidence regarding the effectiveness of networks The grey literature on partnerships and collaboration in health and public health includes descriptive case studies of many different initiatives and interventions aimed at improv ing population health in some way. However, t his collection of descriptive works highlights a fundamental challen ge within this area of research: most partnerships and collaboratives are locally driven, formed to addre ss a specific barrier improve acce ss or implement a specific intervention in their local community. F ew studies of networks in health and public health are generalizable. Significant reasons for this are the unique variations within the public health systems between, and among, counties and states. Consequently, this study aims to build upon established systems and network research to investigate patterns of relationships within the health and public health systems, specifically within stakehol der networks working to overcom e the segmented E I system with the aim to create new generalizable knowledge. Network Performance Although the two bodies of network literature, social networks and organizational networks, inform network performance there is additional complexity that must be taken into account when measuring performance in health and public health. A s discussed in Chapter II, early intervention (EI) efforts to offer physical, occupational, and speech therapies to young children with developmental delays or disabilities have been shown to drastically improve fine and gross motor development, as well as readiness for school, which then can have positive impacts on the & Ziglar, 2002; Haskins, 1989). However, for a stakeholder

PAGE 90

72 network working within the EI system, how can one best measure performance? A count of the number of children receiving EI could be one measure, but that would hardly capture the complexity of the situation. Who is to say that the netw ork was directly responsible for each child receiving early intervention? Further, being referred to EI may not guarantee that the child will benefit if other parts of the system do not function properly. Given the network goals, how can one measure th e impact of EI to the lives of children? If the stakeholder network were life, is that not of tremendous value to that family? Does reducing the number of barriers faced by the families working within the segment ed system count as success, even if the system is still left partially segmented? Would creating awareness about the importance of standardized screening count as success, even if the increased awareness was not directly tied to children being referred? A ll of these questions highlight the tremendous difficulty in which a statistical count can in no way begin to identify collaboration success or failure. It is unlikely a single approach to measure the performance of such a network would be agreed upon in the literature. In fact, some scholars argue that collaborations should not be identified as successful or not successful unless key indicators point in that direction over time and in varying context s (Bingham, 2003). Review of the network literature sh ows that scholars are focusing upon the three structural levels of analysis discussed above node, dyad and network, but when focusing on evaluation of performance, some scholars also measure network performance by looking to community outcomes For examp le Provan and Milward ( 1995 ) measured changes in the incidence of their problem in question and aggregated indicators of client wellbeing. However, Roussos and Fawcett (2000) lament difficulties with this

PAGE 91

73 level outcomes take longer than the Consideration of performance literature outside the realm of networks clarifies the definition of performance measurement Performance measurement is defined in business journal s p. 1229 ). Thus a performance measure for the stakeholder networks is a metric or set of metrics that quantify the efficiency and/or effectiveness of The General Accounting Office (GAO 2012 ) defines performance measur ement as: the ongoing monitoring and reporting of program accomplishments, particularly progress towards pre established goals. .. Performance measures may address the type or level of program activities conducted (process), the direct products and servi ces delivered by a program (outputs), and/or the results of those produ cts and services (outcomes). A program may be any activity, project, function, or policy that has an identifiable purpose or set of objectives ( p 3) Thus applying the two definition s p erformance in this study focuses on the most fundamental question: has the stakeholder network achieved the objectives it was formed to achieve addressing both process and outcome measures. Collaboration researcher s often measure the effectiveness as pe rceived by its members (Thomson et al. 2008 a ) ; while some network scholars have also asked the & Milward, 1995 ). Responding to the challe nges and warning in the literature on measuring performance, as described in the next chapter on Methodology.

PAGE 92

74 Conclusion M any of the many n etwork attributes and characteristics addressed throughout this literature review are not apparent to the organizations within a network because they evolve over time influenced not only from the decisions of that actor, but also from the decisions and perceptions of all the other actors within the network. The structures co nstitute social realities of which the social actors [thus] network research has an emancipatory potential in that it can inform actors of non obvious constraints and opportunities inherent in patterns Kilduff & Tsai, 2003, p. 23). Thus answe ring the research questions posed in this study requires a structured and robust methodology that draws upon network analysis instruments, tools and techniques, as detailed in the following chapter.

PAGE 93

75 CHAPTER IV IV. METHODOLOGY Research Design Overview This chapter provides a detailed description of the design and procedures used in the study. The methods have a sound epistemological grounding in system and network theories, and the procedures are consistent with the network approach Aiming to understand how patterns of interaction, and other attributes and characteristics of community based stakeholder ne tworks affect their performance this study utilize d a mixed method, most different cross case research design The research design has two phases. Based on the methodology of the network approach Phase 1 focuse d on the characteristics and attributes of five stakeholder networks. Whole network and organizational/dyad ic variables were examined through social network analysis (SNA), and analyzed with performance measures based on stakeholder interviews, self assessments and local early intervention assessments. Within case analyses was conducted for each network, and then the findings were further investigated through cross case analysis to identify whether patterns between independent variables and performance were observed. Phase 2 used content analysis to investigate the patterns that emerged fr om the cross case analysis, to better understand how network attributes and characteristics affect network performance. Network findings informed the coding process and the construction of categories. Specifically, content analysis was used to better und erstand

PAGE 94

76 an emergent distinction in the relationship between key network actors and network performance. Each network was treated as a bounded case study, thus allowing for systematic comparison (Yin, 2003) George and Bennett (2005) growing consensus that the strongest means of drawing inferences from case studies is the use of a combination of within case analysis and cross case comparison within a single case study (p. 18) As the purpose of the study is to ide ntify patterns in the attributes and characteristics of networks, relative to their place on a performance continuum, the cross case method with a most different design is a strong choice. The most different cross case research design deliberately seeks t o c ompare cases that are contrasting or different from each other in order to find similar processes or outcomes across cases (Przeworski & Teune, 1970 ). The most different design approach can indicate the robustness of a relationship that is observed bet ween the independent variables and the dependent variable Thus, t he approach lends str ength to the finding s since such an observed relationship would be consistent across contrasting cases (Faur e, 1994 ) This case oriented approach thus emphasizes diversity in the selection of cases (George & Bennett, 2005). Focusing on the dynamic interactions of public health agencies with public and private organizations that affect health, the five stakeholder groups are goal directed networks, worki ng to ensure standardized developmental screening, referral and follow up for children in their Colorado communities. The stakeholder networks were mobilized to overcome the segmented system of early intervention services in their local

PAGE 95

77 community. The ne tworks were purposefully selected through key informant meetings to emphasize diversity in the selection of cases Social network analy sis can be used to measure variables through confirmed and unconfirmed ties. Interorganizational relationships are not a one way endeavor, and analytical methods, such as network analysis capture that complexity. Identifying the Data Needed This study investigated both whole network variables and organizational level variables, as the independent variables, in the context of network performance, the dependant variable. This study first focused on the impact of whole network variables, including density, centralization and trust, on network performanc e, in an exploratory approach question focused on t he impact of organizational level variables, including relative connectivity, brokerage and strength of ties, on network performance. Network Performance Data An understanding of network p erformance was required for investigation into both research questions The stakeholder networks aimed to ensure standardized developmental screening, referral and follow up for children in their community, with the potential 2012 para exact contribution toward this goal and long term outcome d ata on three separate performance assessments were collected to triangulate performance for each of the stakeholder networks: an external assessment by ABCD Colorado leadership team; network member self assessment ; and an early intervention assessment The three

PAGE 96

78 separate measurements were used to enhance construct validity and reliability of the performance measure (Yin, 1994) The rationale and identification of data needed to Variables and Measurements section later in this chapter. External Assessment It is important to understand how the ne tworks were performing from the perspective of an actor outside of the network, but with an intimate knowledge of the local EI systems and necessary efforts towards systems improvement. The Assuring Better Child health Development (ABCD) leadership team fit that bill, and developed a set of deliverables to help inform stakeholder efforts and the measure their progress The ABCD Model Community guide is a process measure of performance, informed by ience about standardized developmental screening and the referral and follow through process that should occur once children are identified The guide also takes the factors that need to be considered when multiple community organizations are performing and advocating for standardized screenings into account (ABCD, 2011). Given the difficulties in accomplishing and measuring systems building related efforts, this external assessment provides an important perspective of network performance. The six key a ctivities and supporting action steps identified in the guidelines are detailed in the Variables and Measurements section of this chapter. Network Self Assessment perception of success is advocated in the literature as another importa nt perceptive of performance. Asking each member of each stakeholder network to rate how successful the network focus been for children in their community captures this important perspective.

PAGE 97

79 Early Intervention Assessment It is also important to capture an objective assessment in the triangulated performance measure. There is no available data on screening rates at the community level; however, an indicator that children with developmental delays or disabili ties are being identified early, and the referral and follow up processes between agencies are working to ensure that children are receiving early intervention services was needed. The researcher looked to Early I ntervention Colorado for such an indicator. The degree to which each loc al community meets the EI targets where the stakeholder networks operates was calculated from the local EI reporting data and was the third measure used to triangulate network performance. After identifying the performance data needed, this investigator next focused on the characteristics and attributes of the networks This required relational data for each of the stakeholder networks, meaning that all organizations in each networks needed to provide information about their relationships with all of th e other o rganizations within the network Whole Network Data sought to better understand the impact of whole network characteristics and attributes on performance asking: d o community based stakeholder networks operat ing within the same context exhibit commonalities in whole network level attributes and characteristics relative to their performance? This is an area of considerable interest in health and public health systems research, but is underdeveloped in the network literature. Informed by the literature, an exploratory approach was selected. N etwork density was identified as one of the whole network independent variables with the aim to investig ate whether there was a relationship

PAGE 98

80 between the overall densi ty of ties in stakeholder networks and network performance Further, it was important to understand the extent to which organizations were more or less connected in the overall network structure. D egree centralization was selected as a measure to capture the way the network was structured and how centralized or decentralized it was, with the aim to investigate whether the re was a relationship between network centralization and network performance ( Provan & Lemaire, 2012). These whole network measures and respective equations are detailed later in the chapter. Organizational and Dyadic Data any relationship s exist ed between attributes and characteristics at the organizational level and network performan ce. Thus data at the organizational level were identified, collected and analyzed. Based up on the literature, the organizational level data that capture network : degree centrality, relative connectivity, closeness centrality, betweenness centrality and brokerage relationships (Provan & Lemaire 2012) These variables are described in detail in the Variables and Measurements section of this chapter. Network literature also Th us th e type of relationships between primary care providers and public health and early intervention organizations is important For this area of study, it was important that the relational dat a not simply indicate an awareness of the organizations or frequency of contact; rather, it needed to capture the intentional efforts of organizations to enhance the capacity of the primary care providers to implement and carry out standardized screening a nd strengthen the

PAGE 99

81 follow through process. To this end, the measure of coordinated ties was selected as an organizational variable and is described in the measurement section. The Phase 2 content analysis further explore d emergent fin dings from the netwo rk analyses. Thi s mixed method, cross case research design is summarized in Table IV. 1 ; with the participants, variables and analysis detailed in the following sections Table IV 1 Summary of Research Design. Variables Data Collection Method Source/ Sample Purpose Analysis Phase 1 Network Performance (DV) External Assess EI Assessment Self Assessment Whole Network Characteristics (IV) Density Deg. Centralization Trust Organizational Characteristics (IV) Matrix Reports Survey Survey Survey Survey ABCD Team EI Colorado Survey Sample: All members of five purposefull y selected stakeholder networks. To measure and analyze relationships between whole network organizational / attributes and characteristics and performance N etwork analysis UCINET; W ithin case analysis; Cross case analysis Centrality Measures Rel. Connectivity Brokerage Relationship Type Survey Survey Survey Survey Survey Sample: All members of five purposefully selected stakeholder networks . Phase 2 Focus of Organizational Organization websites Mission, vision and value statements of all organizations within all networks To garner in depth data about emergent patterns from analysis. Content Analysis; Cross Case Analysis Data were collected via online surveys from organizations from all five stakeholder networks detailed in the following sections.

PAGE 100

82 Participants and Site s Focusing on the dynamic interactions of public health agencies with public and private organizations that affect public health, the population of this study includes five stakeholder networks working to ensure standardized developmental screening, referral and follow up for children in their communities. As detailed, t hese networks were mobilized to overcome the segmented system of identification and delivery for early intervention services in Colorado Selection of Stakeholder Networks The five stakehold er networks included in the study were strategically selected based upon key informant meetings with the Colorado Assuring Better Child health Development (ABCD) lead team. This investigator employed purposeful sampling of stakeholder networks guided by t heory and specific criteria. Creswell (2007, p. 118) stakeholder network was se lected from specific criteria to ensure it is consistent with network theories, and to ensure a broad spectrum of performance in which to be able to compare and contrast across cases. The first five criteria were (2007) five characteristics of networks : 1. a relatively stable horizontal articulation of interdependent, but operationally autonomous actors; 2. who interact through negotiations;

PAGE 101

83 3. which take place within a common regulative, normative, cognitive framework; 4. t hat is self regulating within limits set by external agencies; and 5. (p. 6) Additionally, the following criteria for the selection of stakeholder networks were added for this study to ensure similaritie s among networks, except on the dependent variable Each network needed to : 6. have a core vision to improve child developmental heath, either directly or through its social or economic determinants,; 7. have formed as part of the ABCD initiative because no one organization felt that they could address the objectives on their own; 8. have been in existence for at least one year; and 9. be working to ensure standardized developmental screening, referral and follow up for children in their Colorado community. Comparing networks is a difficult task, as they often vary tremendously in their characteristics a n d attributes. Because they form to overcome specific challenges or problem or meet the needs of a specific community or system, they often vary in purpose, scope, me mbership and size Therefore it was important that these networks developed for the same purpose in the same state, and with relatively levels of resource munificence. The first nine criteria narrowed down the number of possible stakeholder networks to 35. The five stakeholder networks in the study were then selected from these 35 based upon an external assessment by the Colorado ABCD lead team in a most

PAGE 102

84 different research desi gn (Przeworski & Teune, 1970); selecting two stakeholder networks from states with the highest assessments from ABCD, two stakeholder networks with lower assessments from ABCD and one stakeholder network in between. This final criterion helped to ensure that the five networks exemplified a range of performance. This informant team has been leading the ABCD initiative in C olorado since 2007, providing technical assistance and support to community based stakeholder groups working to increase standardized developmental screening and improve th e syst em of referral and follow u p for children in Colorado. The ABCD team know s the networks and communities very well, but was able to provide an objective assessment because it is not a part of any one network. As such, the team developed an assessment matri x, based upon the six key activities determined by ABCD to be essential in successfully working to ensure standardized developmental screening, ref erral and follow up for childre n. A total of five stakeholder networks were selected, since five is the maxi mum number of cases recommended in a single qualitative comparative case study (Cresswell, 2007). The key activities are summarized in T able IV. 2 with the full assessment criteria detailed in Appendix 2. Table IV 2 Assessed Stakeholder Activities Network activities central to goal achievement 1.D eveloping community partnerships; 2. C reating a strong foundation; 3.C arrying out well defined and organized physician outreach ; 4. C learly defining the referral and follow up process ; 5.M aking parent education a priority ; and 6.S etting up trainings

PAGE 103

85 The stakeholder networks in this study operate in five counties in Colorado. The specific stakeholder network names and communities are confidential in this study, although t he counties range in population from 14,548 people to 636,963 people, and 5,330 households to 227,151 households (US Census, 2011 estimate). As diverse as their population size, the land area in square miles also ranges considerably, from 798 square miles to 2,386.10 sq uare miles. Their population densities range from 8 people per square mile to 717 people per square mile while the median household annual income s (2006 2010) range from $40,699 to $60,433 per household (US Census, 2010). Stakeholder Network Members T o address the needs of their communities, each network includes a mix of government, private and nonprofit organizations operating within the health care system, public health system, education system, and system of human services, including the early intervention system in their community. Each of the five networks include s the following types of organizations : Primary Care Providers: Anywhere from one to eight primary care pediatricians or family medicine physicians are part of the stakeholder networks working to ensure standardized developmental screening, referral and follow up for children in their community in Colorado. These primary care physicians have not tradit ionally been a part of the EI system in Colorado, given that EI services are administered by the CDHS, and provided by CCB therapists or contactors ; however there is strong growing consensus about the importance of pediatric and primary care physicians do ing standardized screening with tools such as the Ages and Stages Questionnaire (ASQ) to identify children with developmental delays, or at risk of developmental delays. The published

PAGE 104

86 mission statements of the providers vary, but most detail a primary goa l of providing the highest quality and most comprehensive pediatric health care available in a private practice setting. Central to the identification phase of EI, these primary care physicians bring the perspective of the health care sy stem to the stakeh older network; however, their primary focus is on providing quality individual care to their patients. Local Public Health Agencies ( PH s): The Colorado Department of Public Health (CDPHE 2012). PH agencies provide a broad spectrum of services to address public health and safety issues. They offer services and programs mandated by state s tatute, federal/state funding, as well as those that reflect the unique needs of their community. As such, PH agencies are delivering services and programs to meet 2011 2015 Colorado MCH Priorities identified in the Title V federal funding. Further, The C hild, Adolescent, and School Health (CASH) unit at CDPHE also works with local public health agencies to develop and implement strategies to improve the health of children and adolescents in their communities (CDPHE, 2012a ). The local PH agency within each community is a part of the stakeholder network. primary focus in early intervention is on capacity building, and strengthening partnerships and collaboration with organizat ions and resources that exist in the community to reach its EI related goals. Community Centered Boards : The Colorado Department of Human Services (CDHS), Division for Developme ntal Disabilities (DDD) is the s

PAGE 105

87 leadership for the direction, funding, and operation of services to persons with development al disabilities within Colorado 2012 para 4 ) CDHS implements Part C of IDEA for the State of Colorado by authorizing Community Centered Boards (CCBs) to be responsible for the provision of early intervention services, along with other support and services for individuals with developmental disabilities. CCBs, in partnership with private service providers and therapists, manage and deliver services to individuals with develop mental disabilities and their families throughout the state. The twenty CCBs serve approximately 11,000 (adults and children) and their families in every county across the state ( CDHS, 2012 ). As the EI service provider t he local CCB is also a part of the stakeh older network in each community Focused on providing services, the provision of quality care and services to individuals with disabilities or developmental delays. Early Childhood Councils: According to authorizing leg islation, the role of accessibility, capacity and quality of early childhood services for children and families para. 1). Early childhood services are defined by the state legislation as including: early learning; family support and parent education; social, emotional, and mental health; and health (CDE, 2012). Early Childhood Councils across Colorado bring together partners from each of social/emotional, and parenting opportunities available to all young children and their para. 3).

PAGE 106

88 Educational Organizations that are a part of the stake holder network can include School Districts, which oversee the local public education system within their district; Local Head Start Organizations, which provide early learning opportunities for children from families with low incomes ; and Boards of Cooper ative Educational Services (BOCES), which are an extension of the local school districts. BOCES provide services to children, families and school staff where economies of scale are achieved by delivering across the district rather than through individual schools (BOCES, 2012). Child Care Providers and Early Learning Centers/Preschools are also part of the stakehol der network in each community The behavioral health or community mental health center in the community is part of each stakeholder network as well. It aim s mental health and substance abuse treatment, housing, education and employment (MHCD, 2012 para. 3 ). These o rganizations thus bring a mental health perspective. Other organizations that may be included in the stakeholder networks include health care therapists or home visitation programs, faith based organizations and nonprofits operating the local communities. Census of Ties The population in this study is comprised of all organizations that are a part of the five stakeholder networks. This approach is consistent with network theories as it takes ties in a population of actors, & Riddle, 2005, ch. 1), meaning that the data will contain the entire population of relationships bounded within each collaborative.

PAGE 107

89 Human Subjects Review Prior to the start of recruitment and data collection, this inves tigation was reviewed and approved by the Colorado Multiple Institutional Review Board. As per the Colorado Multiple Institutional Review Board (COMIRB) policies and University of Colorado (CU) investigator responsibilities, the investigator completed the Collaborative Institutional Training Initiative (CITI) Human Subjects Protection Training Course, and submitted the research proposal to COMIRB for review and acceptance before recruitment of data co llection Although the research subjects involved in the project are living, this investigator determined that the project could be appropriately reviewed as non human subject research or be granted exemption under Category B, since the research involved the use of survey procedures focused on organizations (CO M I RB, 2012). The application described the research questions, purpose and methodology, recr uitment, populations, and analysis; declared funding from the National Coordinating Center for Public Health Services and Systems Research and Robert Wood Johnson Foundation; and declared that no conflict of interest existed. The full application is attached in Appendix 3. The investigatio n was approved by COMIRB as non human subject research. Recruitment of Subject Population Upon approval by COMIRB and completio n of the network selection, the investigator presented the details and implications of the study to the ABCD Colorado lead. Information about the study was also provided to the liaisons for each community in which the five stakeholder networks operated. The ABCD liaisons, in turn, emailed the stakeholder members to convey this information and let them know to expect an

PAGE 108

90 invitation from the investigator This was an important series of steps to create awareness unique position, and enable entre to the stakeholder networks. Data Collection The network survey was the primary data collection vehicle in this study. The b ) PARTNER Tool, a validated instrument for network analysis. The comprehensive research that went into the development of the tool, and its extensive use in the field minimized the need for pilot testing. Adaptations and additions to the survey were made to tailor the lan guage to the research setting The consent, re cruitment material and survey questions were presented to the ABCD lead te am for information and feedback, and m odifications were made as necessary The investigator worked with ABCD staff and community liaisons to get contact information for a represent ative from each organization in each of the five networks. An email invitation to participate in the network survey was then sent out to all identified organizational representatives in the network s. The email contained furt her information about the stud y; the consent form ; investigator contact details ; and a unique login and password to the online PARTNER survey shown in Figure IV.1 The specific language in the invitation is attached in Appendix 4. The investigator followed up via email in three to f our days, and again as necessary.

PAGE 109

91 Source: Partnertool.net Figure IV 1 Screenshot of PARTNERTool Survey Login Respondents were asked to complete the surveys as a representative of their organization When respondents logged in, consent was required before the questions were displayed. T he first set of survey questions then requested information about the organization and its role in the network, as well as the existence (or non existence ) of relationships with all other organizations in the network. Building upon that list, the network questions asked each respondent about the types and frequency of interaction between their organizations and all the other organizations in thei r list These types of questions produce data that can be analyzed to determine strength or weakness of ties, connectedness, and measures of centrality, as well as determining brokerage roles within the network. Other questions asked about the level of commitment of organizations to mission, or the reliability of organizations in their work, to determine trust and value at the node, dyad and network level s Questions also ask ed respondents for their organization's perception of network out comes and success, capturing the second of

PAGE 110

92 the three dimension s of network performance in this study T he full survey instrument is attached in Appendix 5. Privacy, Confidentiality and Minimal Risk Information about the organizations involved in the stak eholder networks in each of the five communities was collected through the PARTNER Tool at: http://www.partnertool.net/. Data were maintained in the PARTNER Tool program for the duration of the study. Survey data were also ex ported from the PARTNER Tool, and maintained on an encrypted laptop, in password protected files. T his investigation met the regulatory definition of minimal risk because the probability and magnitude of harm or discomfort anticipated in the research were not greater in and of themselves than those ordinarily encountered in daily working life (COMIRB, 2012) The only research activities that involve d human subjects were through the use of survey procedures. There was no physical or psychological examination test or intervention. The survey wa s completely voluntary, and since it was c ompleted online it could be started and finished at anytime that best suit ed their schedule. Further, a ll findings are presented as an aggregate score for an organization, or at the network level. Variables and Measures Network Performance Dependent Variable Investigati on into both research questions require s an understanding of network performance, the dependent variable in this investigation. T hree separate measures were collected and analyzed to triangulate performance scores for each of the stakeholder networks: external stakeholder assessments, self assessment and local early intervention

PAGE 111

93 assessment. All three measures were summed with equal weight to determine each n External Assessment The Assuring Better Child health Development (ABCD) leadership team developed a step by step guide of tangible deliverables to help inform community efforts and the measure their progress with this work. Th e six key activities and supporting action steps towards achieving a model ABCD community are summarized below: 1. Developing community partnerships. Deliverables: partners identified and roles defined and agreed upon. 2. Creating a strong foundation. Deliverabl es: comparison of actual system with the ideal system in their community; collaboration; safety net strategies; measurement and evaluation; and funding. 3. Carrying out well defined physician outreach. Deliverables: review and verification of local health car e provider data; prioritization of health care provider outreach; development of a plan to initiate and/or strengthen and sustain outreach efforts; health care provider messaging; and use data to support physician outreach and technical assistance. 4. Clearly defining referral and follow through process. Deliverables: review of state referral algorithm; agreement on the local process; partners understand their roles and relationships within the algorithm; partners understand where to refer children partners understand where to refer child who do not flag but still may need extra help; partners are aware of and understand Early Intervention Colorado referral process and referral form; partners are aware of and understand the Child

PAGE 112

94 Find referral proce ss and referral form; partners have clear expectations from referral process and understand referral status update form; partners use data to inform and support community referral efforts. 5. Making parent education a priority. Deliverables: parent education group active and engaged; three priorities for parent education identifies and pursued. 6. Setting up trainings. Deliverables: training delivered on developmental screening tools, child development, and referral proce sses as appropriate (ABCD, 2011 ). The ABCD Colorado leadership team assessed the performance of the stakeholder networks according to the guidelines and deliverables described above. Each group was given an overall scored based upon points earned for each of the six steps and deliverables. Netwo rk Self Assessment efforts in the survey, asked: f this focus been for children in your community ? Respondents were a sked to answer on a to 5= Completely Successful and averaged for a self assessment score. Early Intervention Colorado Assessment Part C of IDEA requires that the lead agency in each state monitor and supervise all early intervention services and programs with in the state (EI Colorado, 2011 ). Although it is difficult to tie child outcomes (su ch as the improved acquisition and use of knowledge of skills) to network performance, it is important to incorporate an element of

PAGE 113

95 the system performance, since the network aims to improve the system. Thus, a measure of the system output is the third mea sure used to triangulate network performance. The Colorado Department of Human Services is the lead agency for EI in Colorado, and is required to report performance data from local programs in relation to the state targets as identified in the Colorado Sta te Performance Plan: Federal IDEA, Part C Services. Indicator 6 in the Colorado State Performance Plan: Federal IDEA, Part C Services for 2010 assesses the percentage of children from birth to age 3 with Individualized Family Service Plans (IFSP) in the EI system, as compared with the stat ewide target (EI Colorado, 2012 ). The data are available on the Early Intervention Colorado website by early intervention program under: Monitoring Reports, Plans of Correction, Public Performance and Determinations. The degree to which each local community meets its targets where th e stakeholder networks operate was calculated as the third dimension for network performance Network Demographics Stakeholder network demographics were also measured, including : size, measured by the number of organizations in each network; age: measured by the number of years the organizations have been working together in a focused effort to ensure standardized developmental screening, referral and follow up for children in the ir community; and relative county size, measured by the population estimates from the 201 0 census. Whole Network Independent Variables Density Density is a core measurement of network structure, which measures the number of total ties in the network. The density of a network is the total number of ties divided

PAGE 114

96 by the total number of possible ties (Borgatti et al. 2002) This measurement helps to illuminate the overall connectedness among organizations in the network, from a strictly numbers point of view. All organizations in each stakeholder net work where asked: f rom the list, please select the organizations/programs with which you have an established relationshi p (either formal or informal) (Survey Question 10). R elationship s were determined based upon the respondent select ions of the other organizations A w hole network measure of density D was calculated for each stakeholder network where l is the actual number of ties and is divided by the total possible ties ( where n is the number of actors in the network) Density (Scott, 2000): This measurement helps to illuminate the overall connectedness among organizations in the network, from a strictly numbers point of view. This measurement allows for an easy understa nding of the connectedness of a network; however, it should be used cautiously when comparing networks of different sizes given that the measurement is sensitive to the number of actors (Scott, 2000). Degree Centralization Degree centralization expresses h ow tightly the network is organized around its most central point (Freeman 1979). This measure complements network density by measuring how the network is organized, and whether it has a centralized structure. Higher centralization scores indicate that the network is more centralized, meaning that the network is organized around one or a few highly central organizations Lower centralization scores indicate that t he network is less centralized meaning that the

PAGE 115

97 network may have ties spread more evenly among actors (Scott, 2000 ). UCINET computes the network degree centralization of a network where M ax is the maximum possible value and is the degree centrality of node and n is the number of actors in the network Degree Centralization (Freem an, 1979) : Network Trust As detailed in literature review, organizational network and collaboration literature emphasize the importance of trust for effective interorganizational relationships and networks (Uzzi, 1996 ; Bryson, Crosby & Stone, 2006; Thomson at el., 2008 b ). To this end, trust was analyzed at the whole network level and compared across cases to investigate whether a positive relationship between trust and network performance existed. Requiring a measure of whole network trust, this study employs Varda et (2008) three dimensions of trust (reliability, share a common vision and openness to discussion) Whole network measurement of trust was measured by the average trust score of all three dimensions of trust in the network. Specifically, organization s in each stakeholder network were asked : How reliable is the organization/program (in the context of standardized developmental screening, referral and follow up for children in your community)? (Survey Question 16); To what extent does the organization /program share a common vision of standardized developmental screening, referral and follow up fo r children in your community? (Survey Question 17); and

PAGE 116

98 How open to discussion is the organization/program (in the context of standardized developmental screening, referral and follow up for children in your community)? (Survey Question 18). Answers were requested on a scale of 1 to 4 (1= Not at all ; 2= a small amount ; 3= a fair amo unt ; 4= a great deal ). The sum of the three responses from each organization in each of the stakeholder network s were then divided by the highest possible trust between all organizations (12 points) Organizational Independent Variables attributes and characteristics, measured at the organization/dyad level, as opposed to the whole network level. The variables are informed by an established area of network theory that asserts correlations between organizational level variables and overall network outcomes Degree Centrality Degree centrality is the extent into which an act or in a network is directly connected to the other actors in the network (Freeman, 1979). Central organizations have more connections to others in the network. The degree centrality is count of the degree for a given point ( ) on the network graph. where 1 i f and are connected and 0 if they are not connected This is divided by the maximum degree centrality, w here can be adjacent to other points in the graph (Freeman, 1979).

PAGE 117

99 Degree Centrality (Freeman, 1979) : Based upon the relationships respondents designated in the network survey, the measure was calculated for all organizations in each of the five stakeholder networks. Betwee n ness Centrality Betweenness centrality reflects the extent to which an organization is on the shortest path between two other organizations, and can thus moderate and influence the information between them Where is the number of geodesics linking two points, that contain p k the betweenness centrality of an actor is a sum of it partial betweenness values for all unordered pairs of actors where i j k (Freeman, 1979). Betweenness is therefore a measure of the number of times a point occurs on a geodesic. The normalized betweenness centrality is the betweenness divided by the maximum possible betweenness expressed as a percentage (Borgatti et al 2002). Betweenness Centrality (Freeman, 1979) : Betweenness centrality is used to investigate whether an organization takes on the role of a gatekeeper within the network, calculated by the extent that an organization is a tie between other organizations. For example, if the same type of organization has high measu res of betweenness centrality in the high performing networks in this study, this may provide an important clue worth further investigation and consideration.

PAGE 118

100 Brokerage Brokerage occurs when, in a group of three or more organizations, one organization requ ires another organization to reach a third organization. (More specifically, given A, B and C, brokerage occurs when A has a tie to B, and B has a tie to C, but A has no tie to C. A needs B to reach C, and B is theref ore a broker) (Borgatti et al. 2002). UCINET calculates the broke rage measure proposed by Gould and Fernandez (1989) by counting the number of times that each organization is a broker in all group possibilities within the network. Brokerage was calculated for each organization based upon a re sponse of monthly or more to the question: At this point in time, how frequently does your organization/prog ram work with this organization on issues related to standardized developmental screening, referral and follow up for children in your community? ( S urvey Question 11 ). This was calculated for each organization within each stakeholder network. Each of these variables were measured at the organizational/dyadic level and compared across cases to investigate whether patterns between organizational level variables and network performance existed. Relative Connectivity Building upon centrality measure s relative connectivity focuses on both the number and quality of connections between organizations in each networ k. The measure is a function of: a ) each o rgani b ) the level of trust that other o rganizations within the network have for them ; and c ) how much the organization value s th e relat ionships with the other organizations. These three measurements are described below:

PAGE 119

101 a) Degree C entrality. Degree centrality is measured by the number of direct ties that an organization has within a network. Key to this study, this measurement also allows investigation into whether an organization is in a central position, or is connected along the periphery of a network. b) Trust. The whole network variable of trust described in the previous section is an aggregate of the trust measures for each organization in each network. Thus, the three dimensions of trust measured at the organizatio nal level can be analyzed at the organizational level and whole network level. For the trust measure, organizations and openness to discussion on a scale of one t o four: reliability : this organization is reliable in terms of following through on commitments; mission congruence : this organization shares a common vision of the end goal of what working together should accomplish; openness to discussion : this organi zation is willing to engage in frank, open and civil discussion (especially when disagreement exists) (Varda, 2012a). These results were summed and a mean trust score was calculated for each organization in each stakeholder network. c) Value. These dimensi ons of value are measured in this study and make up the third element of relative connectivity. Organizati ons were also asked to rate each resource contribution on a scale of one to four :

PAGE 120

10 2 power/influence : the organization holds a prominent position in the community be being powerful, having influence, success as a change agent, and showing leadership; level of involvement : the organization is strongly committed and active in the partnersh ip and gets things done. contributing resources : the organization brings resources to the partnership like funding, information, or other resources (Varda, 2012 a ). These results were summed and a mean value score was calculated for each organization in e ach stakeholder network. Relative connectivity was the sum of an degree centrally; mean trust score ; and mean value score relative to the other organizations in the network. One organization in each network is attributed the highe st relative connectivity of 100 percent Relationship Strength Connectivity of the primary care physicians was conceptualized by the type of activities for which they maintain relationships with other organizations in the network. Members of each stakeholde r network were asked not only about their relationships and frequency of contact with the other organizations in the network but also about the activities that the relationships entailed. All organizations were asked: w hat kinds of activities does your o entail (in the context of standardized developmental screening, referral and follow up for children in your community)? (Survey Question 12). Respondents answered on a rating scale: None (c oded 0);

PAGE 121

103 Cooperative Activities: This may include exchanging information, closing the loop on referrals, attending meetings together, and offering resources to partners (coded 1); Coordinated Activities: Includes cooperative activities in addition to inte ntional efforts to enhance each other's capacity for the mutual benefit of programs (coded 2); and Integrated Activities: In addition to cooperative and coordinated activities, this is the act of using commonalities to create a unified center of knowledge and programming that supports work in related content areas (coded 3). The primary care provider (PC) relationships that entailed coordinated activities were integral measured a nd compared across cases to investigate whether a relationship between these ties and network performance existed. The conceptual and operational definitions of all major var i ables are detailed in Table IV. 3

PAGE 122

Table IV 3 Conceptual and Operational Definitions Variables and Measures Demograpphic Variables Operationalize d Measure Size of the network Number of organizations actively involved in the stakeholder network Age of network Length of time the stakeholder network has existed. County size Total population of county where the stakeholder network operates. Dependent Variables Conceptual Definition Theoretical link Operational Definition Operationalized Measure Performance External assessment Perceived and actual achievement of goals is a measure of network outcome (Gray 1989, 2000 ; Provan & Milward 1995) model community guide as compared to other networks. self assessment Survey responses recorded on scales ranging from 1 (not at all) to 5 (to a great extent)/ coding. Overall, how effective has the stakeholder network been at reaching its ob jectives? Please indicate which of the following outcomes you feel have been achieved as a result of the efforts (selection of stakeholder objectives as many that apply). From your perspective, which of the above describes the most successful outcome of the stakeholder network (selection of one of the collaborative objectives)? What aspects of collaboration contribute to this success? EI assessmen t Percentage of children with I FSPs the written plan for EI services from the local EI Colorado program as compared to state target Early Intervention Colorado local services data for children, newborn through three years of age with IFSP, as compared to state target. 104

PAGE 123

Independent Variables Conceptual Definition Theoretical link Operational Definition Operationalized Measure Whole Network Variables Network structure Density Central ization related to network effectiveness ( Provan & Milward 1995 ) Network structure calculated via SNA through UCINET program index created from two network measurements collected through survey. The number of ties in a network as a percentage, in relation to the total number of possible ties. Degree centralization How similar or dissimilar the organizations are in terms of their number of connections to others. Trust High levels of trust correlated with network and collaboration effectiveness (Huxham & Vangen 2005) Aggregated mean score from organizational level trust data See in Organizational Variables. Organizational Variables Connectivity Degree c entrality Strategic connections can improve the network functionality and outcomes (Burt 1995 ; Freeman, 1979; Granovetter 1973; Krackhardt, 2010 ; Kilduff & Tsai 2003 ; Varda et al ., 2008 ) Network interaction calculated via SNA through UCINET 6 program and PARTNER Tool. The number of connections that organizations have to other members of the networks. Betweenness c entrality between two other organizations. Brokerage The extent to which an organization connects otherwise unconnected organizations in a network. Relative c onnectivity The highest number of quality of connections. Strength of t ies The n umber of ties representing coordinated activities: mutual benefit of programs. Value Power/influence The levels of organizational values can affect effectiveness; not just power as traditionally thought, but also involvement and time contribution (Varda et al ., 2008) Organizational value score calculated from survey responses recorded on scales ranging from 1 (not at all) to 4 (to a great extent) The organization holds a prominent position in the community be being powerful, having influence, success as a change age nt, and showing leadership. Level of involvement The o rganization is strongly committed and active in the stakeholder network and gets things done. Resource contribution The organization brings resources to the stakeholder network like funding, information, or other resources. 105 Table IV.3 (cont.)

PAGE 124

Trust Reliability High levels of trust correlated with network and collaboration effectiveness (Huxham & Vangen 2005); Three dimensions of trust: 'reliability', 'in support of project' and 'open to discussion' (Varda et al ., 2008) Trust score calculated from survey responses recorded on scales ranging from 1 (not at all) to 4 (to a great deal ) This organization is reliable in terms of following through on commitments. In support of project This organization shares a common vision of the end goal of what working together should accomplish. Open to discussion This organization is willing to engage in frank, open and civil discussion (especially when disagreement exists). 106 Table IV.3 (cont.)

PAGE 125

107 Data Analysis Analysis of data collected in Phase 1 is based up on the methodol ogy of network analysis As detailed in the previous chapter network analysis enables the stakeholder networks to be measured and analyzed mathematically, producing calculations of the overall network structure and the attributes and characteristics of the relat ionships embedded within as well as producing a visual map (graph) (Scott, 2000; Wasserman & Faust 19 94). Network analysis provided a powerful and precise method to visually depict a network; and an excellent analytical tool to calculate and map the network attributes and characteristics (Scott 2000). Network Analysis Given the systems approach required to answer the research question, network analysis was undertaken to graph and calculate key attributes and characteristics of the five stakeholder networks. Graphs are useful to present information visually, enabling one to see patterns that may not otherwise be evident. Network matrices also enable the use of complex mathematical algorithms to calculate patterns in attributes and characteristics (Hanneman & Riddle, 20 05). In order to answer the research question s binary measures of relations had to be calculated for each network separately from the survey data. At the most basic level, r espondents from each stakeholder network were asked: f rom the list, please sele ct the organizations/programs with which you have an established relationship (either formal or informal) (Survey Question 10). The responses were assigned numbers to distinguish between the presence or absence of a tie (coded one or zero, respectively). Multiple

PAGE 126

108 category nominal measures of relationships were also collected and analyzed. For example, respondents were asked: a t this point in time, how frequently does your organization/program work with this organization/program on issues related to standa rdized developmental screening, referral and follow up for children in your community? (Survey Question 11). Respondents were asked to scale their relations from the list below to gather information about the strength of the ir ties: Never/We only interact on issues unrelated to standardized developmental screening, referral and follow up; Once a year or less; Every few months; Every month; Every few weeks; or Every week. Further, other questions asked about the kinds of activities that the relationships e ntailed. For these, each relationship was coded on a qualitative scale by its type, rather than its strength. Unlike the true false nature of the binary nominal data, the multiple category nominal measure is multiple choice (Hanneman & Riddle, 2005). No rmative information such as trust and value was also collected, given the importance of these characteristics in the literature Since network analysis is based on a census, rather than a population sample, it is always important to attain a response rate as close to 100% as possible. While no finite limits of survey response rate have been determined in the network research to date, esteemed network scholar Stephen Borgatti (2012) asserted that a response rate of close

PAGE 127

109 to 70% is important to maintain research validity and reliability in network research The stakeholder network response rates in this study were as follows: Alpha: 10 of 12 organizations responded for an 83% response rate; Bravo: 19 or 27 organizations responded for a 70% response rate; Charlie: 8 of 10 organizations responded for an 80% response rate; Delta: 12 of 18 organizations responded for a 67% response rate; and Echo: 10 of 14 organizations responded for a 71% response rate. The PARTNER Tool ( Va rda 2009b), UCINET 6 (Borgatti et al. 2002 ) and NetDraw (Borgatti et al. 2002) are the software programs used to analyze the survey data, and to create the graphical representations of the networks. The survey responses were stored in the PARTNER Tool, since respondents completed the survey online directly through the PART N displayed the raw survey data in Excel files. Preliminary data analysis was conducted using PARTER Tool functions to convert the re lational data into network visualizations, and to complete calculations of network measurements including relative connectivity, trust and value. The binary and value data for each stakeholder network were then manually converted from the PARTNER Tool in t o edgelist text files and im ported into UCINET for further analysis UCINET converted these files into sociomatrices. Entries code a relationship from the organizations in the row to the orga nization in the column By way of example, a matrix of relatio nal data for one of the stakeholder networks is depicted in Figure I V. 2

PAGE 128

110 Organizations 1 2 3 4 5 6 7 8 9 10 11 12 Org 1 0 0 0 0 0 0 0 0 0 0 0 0 Org 2 0 0 0 0 0 0 0 0 0 0 0 0 Org 3 2 1 0 3 2 5 4 4 3 2 3 3 Org 4 0 0 0 0 1 3 2 2 0 0 0 0 Org 5 3 1 3 3 0 3 3 3 3 3 3 3 Org 6 3 0 4 3 3 0 6 4 5 3 3 3 Org 7 4 0 4 4 4 6 0 4 5 4 4 4 Org 8 3 0 4 4 3 5 5 0 5 4 4 4 Org 9 5 1 1 4 5 5 5 5 0 0 4 4 Org 10 0 0 0 0 1 3 3 0 0 0 2 0 Org 11 3 0 4 3 0 4 0 4 3 0 0 0 Org 12 2 1 3 1 2 0 2 3 3 1 3 0 Figure IV 2 Sample Matrix of Stakeholder Network. [ Valued for frequency of relationship ] This matrix depicts the presence or absence of relationships, based upon responses by each organization to survey question 11 (coded: Never/We only interact on issues unrelated to standardized developmental screening, referral and follow up= 0; Once a year or less= 1; Every few months= 2; Every month = 3; Every few weeks = 4; Every week= 5; Every Day= 6. Sociomatrices were cre ated to represent network ties for each of the relational, value and trust questions in each of the stakeholder networks. UCINET was then used to calculate the whole network measurements of density and centralization, as well as organizational measurement s including centrality, betweenness centrality, and brokerage. NetDraw was used to develop additional network graphs and visualizations Finally, a cross case analysis was conducted to look across cases, and led to new generalizations from emergent pattern s between the network attributes and characteristics and network performance. The stakeholder networks were placed along a continuum of performance based upon their aggregate performance measures. The investigator then

PAGE 129

111 examined each of the independent v performance continuum. 2009, p. 5). Further, case study analysis is well suited for research ques (Merriam 2009, p. 19). Further, one of the greatest strengths of case study methodology methods and formal model & Bennett 2005, p.19). Finally, case study methodology allows for multiple cases to be compared, revealing insight into the similarities and differences among cases, enabling generalizations to develop through analysis (Creswell 2007 ). Content Analysis A follow up phase to the network analysis emp loyed a qualitative methodology to help capture the rich detail of the attributes and characteristics being explored. themsel ves and their settings and how inhabitants of these settings make sense of their ( Berg 2004, p. 7). NVivo 9 software was used for the coding process and qualitative analysis. As per qualitative analysis practice, these categories were mutual ly exclusive and conceptually congruent (Merriam, 2009, p 186). Cresswell (2007, p analysis, then reducing the data into themes through a p rocess of coding and condensing

PAGE 130

112 categories and codes for Phase 2 analysis were not developed a priori, but were informed by the findings that emerged from the networ k cross case analysis. The investigator coded the text of the mission, vision and value statements w i th descriptive labels, and process was then repeated again to From these, two aggregate dimensions emerged from the process to illuminate a distinction in organizational t ype of the key network actors Study Validity and Limitations Validity This investigation is a comparative case study of five goal directed stakeholder networks, with the aim to indentify and explore patterns. Each stakeholder network is treated as a distinct case study. The threats to external validity of the study c ould b e a concern given the small N of five organizational networks; however no causal relationships or statistical correlations are being sought The aim of this study to shed light on the research questions posed to help policy makers and program impl ementer s better understand organizational networks through theory based empirical research. Given the mixed method, cross case design, there are low threats to maturation, testing effects, instrumentation and attrition. In an attempt to strengthen construct va lidity and (2012 a ) validated and field tested PARTER survey. However, with the focus on the organization, it is important to recognize that individual representatives completed the survey on behalf of the organizations I t is possible that these individuals might have

PAGE 131

113 been un aware of all ties between their organization and the other organizations in the network ( in the context of standardized developmental screening, referral and follow up for children in their community ). It is also possible that some of the answers were influenced by their own personal biases or beliefs, and do not accurately reflect the quality of ties between their organization and others This is inherent in all organizational network research. The respondents were the individuals actively involved in the network, representing their organization in the ongoing efforts to improve standardized developmental screening, referral and follow up for chil dren in your community so this threat is minimized to some extent. Although significant efforts were made to attain as high a response rate to the network survey as possible to improve the validity of the data not every single organization comp l eted the network survey. In the end, 61 organizations complete d the survey out of a total of 81 organizations This is an overall response rate of 75% with individual network response rates ranging between 67% 83% ; however, it is important that the resp onse rate within each network be as high possible to ensure that there is not an unequal representation of ties that could bias the results. There is a low threat to statistical regression as this research design does not select on extreme scores, does n ot have a second test, and the analysis is mathematically based instead of focusing on predictive relationships. Finally, while there is a medium threat to selection, because the study is based on one type of network: this study investigates stakeholder n etworks in only one specific subsystem of the health and public health care systems: that of early intervention (EI). This threat is partially overcome with the selection of five stakeholder networks. Further, the ultimate purpose is to identify

PAGE 132

114 patterns that emerge, thereby shifting the focus of the findings to only those variables that were common across all sites, further reducing the threat to internal validity. Replicating this research with another set of organizational networks in a different cont ext to explore whether the same trends hold true would help to mitigate the external threat to validity. Limitations It is important to recognize the limitations of any research. Of the four systems approaches detailed in Chapter I (systems organizing, system dynamics, system networks, and systems knowledge), this study is primarily focused on just one: system networks. The findings from the cross case analysis are thus limited to the relationships between the organizations involved in the stakeholder n etworks, and do not take into account all of the relationships within the EI system. Another limitation to the study stems from the inability to directly measure the increase in standardized screening, referral and follow through in each network communit y. The scope of such a performance measurement is well beyond the means of one investigator, and does not exist in secondary data sets at the level required for this investigation. Further, the ultimate aim of the networks is to enable children with deve lopmental disabilities to reach their maximum development potential, an outcome that would be nearly impossible to quantify. This weakness was overcome to some degree by triangulating performance data from multiple sources: external stakeholders; members of the stakeholder networks, and Early Intervention Colorado. Further, data were collected at just one time from the networks. Interorganizational relationships are dynamic and always evolving. The findings would be strengthened with a research design t hat called for multiple points of data collect ion.

PAGE 133

115 Finally, the inability to determine statistical correlation or causal relationships is another area of weakness for this investigation. Any generalizations from the study findings should be treated wit h caution. A larger N would help to overcome this weakness.

PAGE 134

116 CHAPTER V V. FINDINGS This chapter shares findings of the cross case analysis of five stakeholder networks operating within the early intervention system in Colorado which spans the health and public health systems, along with systems of human services and education The research question s that guide this study ask : Do community based stakeholder networks operating within the same cont ext exhibit commonalities in network level attributes and characteristics relative to their performance ? D o the connectivity and structural positions of member organizations affect stakeholder network pe rformance ? T o answer these questions t he first step was to analyze network performance and place the stakeholder network s on a performance continuum. The second step was to analyze each stakeholder network and compare all five cases to identify similarities or differences in demographics, structural and normative characteristics at the whole network level including density degree centralization and trust. The third step in the analysis sought to identify and analyze patterns of similarities and differences at the dyad and organizational levels of a nalysis including degree centrality, relative connectivity, closeness centrality, betweenness centrality and brokerage (IV s ) in the context of stakeholder network performanc e (DV) The fourth step in the analysis sought to explore and interpret the dist inction in organizational type that emerged from the network cross case analysis findings through content analysis of C ross case analysis was conducted utilizing UCINET and NetDraw (Borgatti et al. 2002) and the PARTNER Tool (Varda, 2008 b ) on the data set containing five

PAGE 135

117 stakeholder networks (N n =5); 81 total organizations (N o = 81); 61 respondent organizations (N r = 61); and 483 dyadic ties (N t = 483). Content analysis was conducted utilizing NVivo 9 on t he mission and vision statements of each organization in the five network s Findings are presented in the sections that follow. To ensure confidentiality to the stakeholder network members, the networks have been coded A E using the International Radiotelephony Spelling Alphabet: Alpha, Bravo, Charlie, Delta, Echo in order of performance. Performance Both research questions require an understanding of network performance. Data on t hree separate performance measures detailed in Chapter IV were collected and analyzed to triangulate performance assessments for each of the stakeholder networks Three distinct assessments were used to enhance construct validity and reliabil ity of performance, the dependent vari able in this study (Yin, 1994). The performance assessments for each stakeholder network are summarized below and detailed in Table V. 1 External Assessment The Assuring Better Child health Development (ABCD) Colorado leadership team assessed the performance of the stakeholder networks according to six key activities and deliverables to help inform community efforts and measure progress with their efforts to e nsure standardized screening, referral and follow up for children. The six key activities and subsequent deliverables were described in Chapter IV and are attached in Appendix 2 Each stakeholder net work was given an overall assessment based upon points earned for each of the six steps and deliverables. The raw assessments ranged from

PAGE 136

118 a low of 22 to the highest score of 70 with a mean of 42 and a standard deviation of 18 The highest possible score was 100. To equally weight the three different perfor mance assessments in the triangulated measure of network performance, each assessment is calculated as a percentage of highest possible points. This measure was also used to select the networks for i nclusion in the study in a most different cross case stud y design to deliberately compare cases that differ in their performance range (as assessed by ABCD Colorado) in order to identify patterns of attributes and characteristics relative to their overall performance. Upon case selection, t wo additional performance measures were identified to triangulate the performance measure for this study These include: members perceived performance in a self assessment and local EI data in an assessment of system output, as described in Chapter IV an d summarized below. Members of the stakeholder networks were requested via survey, to s elf assess asked to answer on a rating scale. The answers were weighted from 1= Not Successful to 5= Completely Successful and aver aged for a self assessment score The s elf a ssessment scores ranged from an average of 3 to 3.5. The mean was 3.20, with a standard deviation of 0.23. To enable comparison across cases and ensure equal weight in the triangulated performance measure, the self assessment scores are calculated for each stakeholder network as a percentage of the highest possible score. For example, weighting the responses from 1 to 5, the total score for Alpha was 35. Ten organizations provided

PAGE 137

119 responses, for a possible ma 35/50= 70 percent. For ease of discussion, the decimal point is dropped and the percentage is equated to a score between 1 and 100 meaning that the score of 70 percent becomes 70. Local Early Int ervention Assessment Part C of IDEA requires that the lead agency in each state monitor and supervise all early intervention services and programs within the state (EI Colorado, 2012 ). The degree to which each local community meets the EI targets where the stakeholder networks operate s was calculated from the local EI data described in Chapter IV, which was the third measure used to triangulate network performance. Three of the five communities met or exceeded the target earning a score of 100 The mean EI assessment was 92 with a standard deviation of 10 The three measures of performance are summed into a performance index. Based on this index, the mean performance of the five stakeholder networks in this study was 199 with a standard deviation of 25 A summary of the network performance assessments are presented in Table V.1. Table V 1 Network Performance. Stakeholder Network External Assessment Self Assessment Local EI Assessment Performance Index Alpha 70 70 100 240 Bravo 43 62 100 205 Charlie 43 68 80 190 Delta 22 60 100 182 Echo 34 60 83 177 Mean 42 64 92 199 SD 18 5 10 25

PAGE 138

120 Stakeholder Network Alpha had the highest overall performance of the five stakeholder networks in this study, with a performance score of 240 The stakeholder network was assessed among the highest in the external assessment by the ABCD Colorado leadership team with an assessment of 70 out of a possible 100 (70% = score of 70 ) The self assessment reflected a sense of accomplishment, with an assessment of 35 out of a possible 50 (70% = score of 70 ). The distribution of responses from the self assessment is depicted in Appendix 6. The local early intervention data showed that the birth to age three with Individualized Family Service Plans (IFSP), compared to the statewide target of 2.5 percent (e xceeding 100% = score of 100 ). This indicates that children with developmental delays or disabilities are being identified early, and the referral and follow up processes between agencies are working to ensure that children are receiving early intervention services. The assessment also signified progress against the 2012 ). The stakeholder group has been working together towards standardized developmental screening, referral and follow up for children in their community for approximately 18 months at the time of the survey, and is among the smaller stakeholder networks in this study (4 th out of 5). Stakeholder Network Bravo had the second highest overall performance of the five stakehol der networks in this study, with a performance score of 205 The network's external assessment was 43 out of the highest possible of 100 (43%= score of 43). The self assessment reflected a sense of accomplishment, with an assessment of 56 out of a possi ble 90 (62%= score of 62). The distribution of responses from the self assessment is depicted in Appendix 6. The local early intervention data showed that the target had

PAGE 139

121 been exceeded, with 3.16 percent of children from birth to age three with ISFPs tha n the statistically determined statewide target of 2.5 percent (exceeding 100%= score of 100). (EI Colorado 2012 ). The stakeholder network has been working together for almost 6 y ears, and is the largest network in this study. Stakeholder Network Charlie had a performance score of 190 The network's external assessment was 43 out of the highest possible of 100 (43% = score of 43). The self assessment reflected a sense of accompli shment, with an assessment of 27 out of a possible 40 (68%= score of 68). The local early intervention data showed that the target had not yet been reached, with 1.99 percent of children from birth to age three with ISFPs, which is lower than the statewide target of 2.5 (80%= score of 80) The assessment also reflected a lower percentage of children 0 3 with ISFPs than the previous y ). The stakeholder network has been working together for approximately 3 years, and is among the smallest networks in this study. Stakeholder Network Delta had a performance score of 182 The network's external assessment was 22 out of the highest possible of 100 (22% = score of 22). The self assessment reflected a sense of accomplishment, wit h an assessment of 30 out of a possible 50 ( 60 % = score of 60 ). The local early intervention data showed that the local target had been exceeded, with 2.77 percent of children from birth to age three with ISFPs, compared to the statewide target of 2.5 per cent (exceeding 100%= score of 100). However, t his assessment reflected a lower percentage th (EI Colorado 2012 ). The stakeholder network has been working together for approximately 3 years, and is mid sized among the networks in this study.

PAGE 140

122 Stakeholder Network Echo had the lowest performance of the five networks in this study, with a performance score of 177 The network's external assessment was 34 out of the highest possible of 100 ( 34 %= score of 34 ). The self assessment reflected a sense of accomplishment, with an assessment of 27 out of a possible 45 ( 60 % = score of 60 ). The local early intervention data sho wed that 2.07 percent of children from birth to age three with ISFPs, as compared to the statewide target of 2.50 percent (83%= score of 83) sure, however (EI Colorado, 2012 ). The s takeholder network has been working together for approximately 3 years, and is among the smallest of the networks in this study. The three performance measures for the five stakeholder networks are depicted in Figure V. 1 Figure V 1 Comparison of Stakeholder Network Performance. 0 10 20 30 40 50 60 70 80 90 100 External Assessment Self Assessment Local EI Data Network Performance Network Alpha Network Bravo Network Charlie Network Delta Network Echo

PAGE 141

123 As Figure V.1 illustrates stakeholder network Alpha performed well, achieving the highest s core in the external assessment Stakeholder network Bravo tied with Alpha in the self assessment and local early intervention data, and scored second in overall performance Stakeholder network Charlie was in the middle range of performance, scoring fairly consistent ly across all three measures. The stakeholder networks Delta and Echo are close to one another at the lower end of the performance assessments in this study with one network scoring higher in the external assessment, tied in the self assessment and the other network scoring higher in the local EI data. All three of the measurements reflect an important element of network performance. The external assessment focuses on the processes and steps that are screening and the referral and follow up process that should occur once children are identified, as well as factors that need to be considered formance in improving the identification and referral stages of the EI system. This is weighted equally with the local early intervention assessment, which captures performance of the delivery of EI services. These two measures are weighted equally with performance, a perspective that has also been identified as an important measure in the performance across the whole EI system. The stakeholder networks were then placed along a performance continuum based upon the triangulated performance measure ( equal

PAGE 142

124 local EI assessments ) as illustrated in Figure V. 2 Highest Performance Lowest Performance Figure V 2 Stakeholder Networks along a Performance Continuum. The figure reflects a continuum of performa nce for the five networks in this investigation only. Demographic Analysis Demographic Variables The first research question asks whether community based stakeholder networks operating within the same context exhibit commonalities in network level attributes and characteristics relative to their performance. All five stakeholder networks were selected from the same relational of early intervention for the purposes of holding the network environment constant. All five networks aim to improve th eir local EI system by working towards ensuring standardized developmental screening, referral and follow up for children in their community. Although they formed to meet the unique needs of their community, they are have fairly similar levels o f environmental stability, since EI is legislated at the federal level and managed at the state level, as described in Chapter II. Their resource munificence was also fairly stable, with a consistent level of t echnical

PAGE 143

125 assistance, from ABCD. The demogr aphics of the five stakeholder networks are summarized in Table V. 2 Table V 2 Summary of Network Demographics. Stakeholder Network Performance Size Age County Size Rank Alpha 240 12 1.3 1 Bravo 205 27 6 4 Charlie 190 10 3 5 Delta 182 18 3 2 Echo 177 14 3 3 Mean 199 16.2 3.6 N/A SD 25 6.7 1.7 N/A [Size: number or organizations in a network. Age: number of years the network has been operating in community. County size indicates the population of a county that the network operates in rank order relative to the others: small to largest. Mean scores and standard deviation are listed following the networks]. As indicated in Table V. 2 the five stakeholder netwo rks in this study range in size and age The network sizes range from 10 27 organizations, with a mean of 16.2 and a standard deviation of 6.7 The median size of the five networks was 14 organizations. The age of the networks also ranged from 18 months to 6 years old, with a mean of 3.3 and a standard deviation of 1.7. The median age of the network s at the time of data collection was 3 years old. Patterns of size age, and comparative county size are explored in the cross case analysis Cross Case Analysis: Demographic Variables Beginning with the demographic variables, the cross case analysis examines each variables and performance This exploratory phase of the research investigate s the first research question as there are few theories that correlate whole network measures of stakeholder systems building netwo rks to performance.

PAGE 144

126 Figures V. 3 through V.5 illustrate e network size, network age and relative county size. Network Size Figure V.3 depicts the performance levels of the five networks by size. The performance level of each network in the study has been plotted along the y axis and the size of the networks has been plotted along the x axis. Each point represents the performance and size of a network. Figure V 3 Variation in Network Performance by Size. As the graph illustrates, there was no clear pattern or evidence of a relationship in he highest performing network is one of the smallest in the study, while the next highest performing network is the largest in the study. The peaks and valleys can be observed in the Order 4 polynomial trendline, which has an R squared value of 1. A trendline is most reliable when its R square d value is at or near 1, and thus this is a perfect fit of the curve to the data. The trendline shows the variation in the performance of the networks relative to their size This result is R = 1 0 100 200 300 400 500 600 700 0 5 10 15 20 25 30 Performane Continuum Size of Network, number of organizations Network Performance by Size Bravo Alpha Charlie Delta Echo

PAGE 145

127 important in public health, where collaboration is measured, in part by counting the number of organizations. This is finding is explored in Chapter V1, in the context of EI, health and public health systems. Network Age Figure V. 4 depicts the performance levels of the five networks in the study by the number of year s that the network has been working together. Each point represents the performance and age of a network. Figure V 4 Variation in Network Performance by Age. As the graph depicts, there was no clear pattern or evidence of a relationship in a The youngest network scored lowest on the performance continuum, while the oldest network was next on the performance continuum. The peaks a nd valleys can be observed in the Order 3 polynomial trendline The trendline has an R squared value of 0.99 which is a good fit of the curve to the data R = 0.98 150 170 190 210 230 250 270 0 1 2 3 4 5 6 7 Performance Continuum Age of Network, years Network Performance by Age Bravo Echo Alpha Delta Charlie

PAGE 146

128 Relative County Size (Population) The stakeholder networks in this study operate in five count ies in Colorado, which were each ranked by population (1 5) relative to the others. Figure V. 5 depicts represents the performance the network and the relative population of the county in which it operates. Figure V 5 Variation in Network Performance by Relative County Size. As the graph depicts there was no clear pattern or evidence of a relationship in a net performance relative to its county size. The highest performing network is in one of the smallest counties in the study, while second highest performing network operates within one of the largest counties. The peaks and valleys can be observed in the Order 3 polynomial trendline. The trendline has an R squared value of 0.99, which is a good fit of the curve to the data. Overall, no clear pattern s of performance existed in the cross case comparison of demographic variables. R = 0.9826 140 160 180 200 220 240 260 0 1 2 3 4 5 6 Performance Continuum County Population, smallest to largest Network Performance by Relative County Size Charlie Alpha Echo Delta Bravo

PAGE 147

129 Whole Network Cross Case Analysis Whole Network Variables Based upon responses to the network surveys, sociomatrices were developed and characteristic and att ributes at the whole network levels were calculated Table V. 3 details different characteristics of the whole network that are of importance in addressing the first research question. Table V 3 Summary of Whole Network Characteristics. Stakeholder Network Density Densi ty Level Degree Centralizatio n Centralization Level Trust Alpha 81.8 % High 21.80 Low 76.5 % Bravo 33.4 % Low 55.40 High 67.4 % Charlie 77.8 % Mod High 27.80 Mod Low 76.2 % Delta 57.5 % Mod Low 47.80 Mod High 78.7 % Echo 60.4 % Mod Low 37.20 Mod Low 76.3 % Mean 62.2% 38.00 75.0% SD 19.3 % 13.84 4.4 % [Density: percentage of ties present in relation to the total number of possible ties. Degree centralization: relationship between centrality of all members in network. The lower the centralization score, the more similar the members are in terms of their number of connections (e.g. decentralized). Trust: the percentage of how much members trust one another. Mean and standard deviation are in parenthesis]. To enable comparison, categorical levels we re also determined for the structural network variable s A description of High was used to describe measures that were greater than the mean by more than one standard deviation; Moderately High was used to describe measures greater than the mean by less t han one standard deviation; Moderately Low was used to describe measures less than the m ean by less than one standard deviation; and Low was used to describe measures less than the mean by more than one standard deviation ( Masseurs 2012).

PAGE 148

130 Density The density of the stakeholder networks in this study ranged from 33.4 81.8 percent, with a mean of 62.2 and a standard deviation 19.3. Since density is an index of the degree of connection between actors within the network, it can provide some insight ab out the connectedness of a network, but it is sensitive to the number of organizations. Figure V.6 depicts the performance levels of the five stakeholder networks by their density levels. Figure V 6 Varia tion in Network Performance by Density. As the graph depicts, there was no clear pattern or evidence of a relationship in a most dense with a density level of 81.8 % and Bravo was the le a st dense with a density of 33.4 %. The peaks and valleys can be observed in the Order 3 polynomial trendline. The trendline has an R squared value of 0.97, which is a good fit of the curve to the data R = 0.9858 120 140 160 180 200 220 240 260 0 20 40 60 80 100 Performance Continuum Network Density, percentage Network Performance by Density Alpha Bravo Charlie Delta Echo

PAGE 149

131 While d ensity has been interpreted as a measure of netwo rk integration ( Masseurs 2012) it only captures the number of ties present, not how they are organized within the network. To understand network integration, it is important to also investigate how the network structures are organized with respe ct to the centrality of its members, A whole network centrality measure is the degree of centralization, which is calculates the structural positions of all organizations in a network relative to each other. The measure indicates whether a network is centralized around a single organization, or group of organizations, or whether it is more decentralized Degree Centralizatio n The degree of cen tralization ranges from 21.80 55.4 0 This means that none of the stakeholder networks in this study are centralized around a single organization. The cross case analysis of performance relative to degree centralization is illustrated in Figure V.7. Figure V 7 Variation in Network Performance by Degree of Centralization. R = 0.9825 170 175 180 185 190 195 200 205 210 0 10 20 30 40 50 60 Performance Continuum Network Degree Centralization, percentage Network Performance by Degree Centralization Alpha Charlie Bravo Delta Echo

PAGE 150

132 As the graph depicts, there was no clear pattern or evidence of a relationship in a T he highest performing network, Alpha, is the least centralized network in the study with a low degree of centralization of 21.8 percent, while the network next along the performance continuum, Bravo, had a high degree of centralization of 55.40 percent relative to the study mean The two lower performing stakeholder networks in the study, Delta and Echo had a moderate high and moderate low degree of centralization with 47.80 percent and 37.20 percent respectively relative to the study mean The peaks and valleys can be observed in the Order 2 polynomial trendline. The trendline has an R squared value of 0.95, which is a good fit of the curve to the data. The lack of a linear relationship in a ntralization adds to the inconsistent (and contradictory) research findings on this measure. The implication of the finding will be explored in Chapter VI. This is a whole network aggregate measuremen t based on the centrality of organizations within the n etwork. By itself, it cannot provide insight into the structural position of specific organizations within the networks the focus of research question two The results of centrality and betweenness centrality, which measure structural positions at the i ndividual organizational level are investigated within the context of performance and pr esented later in the chapter. Overall Network Trust Based upon findings in the literature it was expected that there would be higher levels of trust in the higher p erforming stakeholder networks. Each survey respondent in the five networks was asked to rate every other organization that they had a tie with in the

PAGE 151

133 network on three dimensions of trust: reliability, support of mission and openness to discussion in a sc scores were then aggregated and divided by the highest possible rating to determine the percentage tru st in the organizations As summarized in Table V. 3 ranges from 67.40 78.70 percent, with a mean of 75 percent and a standard deviation of only 4.4 percent. With 100 percent trust reflecting that all network members trust all other network members comp letely, t he findings illustrate a key similarity among networks: all five stakeholder networks in the study have high levels of trust among organizations. Since all stakeholder networks in this study have high levels of trust, no pattern in el of performance relative to its overall trust could be ascertained. The findings showed that the relatively lower performing stakeholder networks had very similar levels of trust than the high er performing networks in the study Thus, the study finding s indicate that while trust may be asserted in the literature as a necessary condition for network performance, it is not sufficient. The importance of these findings in EI, health and public health system s are discussed in Chapter VI. Network Graphs Structural characteristics of the whole network are also evident in network graph s, which visual ly represent the networks G raphs illustrate a set of nodes which represent the organization in the network, along with a set of ties that signify how they a re connected. Figure V. 8 illustrates ties in stakeholder network Alpha

PAGE 152

134 Figure V 8 Graph of Stakeholder Network Alpha. The network looks quite dense, with many organizations connected t o many other organizations. This is supported by a density c alculation of 81.8 percent detailed previously The network does not appear to be centered around any single organization and the number of ties that each organization has with other organizati on rang es from 4 11 This observation is supported by its network degree of centralization of 21.80 percent, meaning that the network is decentralized. The ties in this graph are dichotomous (coded as 1 = present or 0= absent) The ties also have a parameter, for Figures V.8 above and V.9 that follows, each organization or program must have worked with another organization/program on issues related to standardized developmental screening, referral and follow up for children in their community at leas t once a month for the existence of a tie. Ties can also be value d with strength of relationship and type of re lationship as they are later in the chapter Node (Organization) Tie (Relationship) Local behavioral health center (BH) Early Childhood ( ECC/ E L ) Local school district /BOCE (ED) Local public health agency (PH) Local primary care provider (PC) Local EI service provider (EI) Parent/Family Other nonprofit

PAGE 153

135 The organizations themselves als o have attributes: in Figure V.8 each organization is coded by ty pe (behavioral health; early childhood; education; public health; health care; or early intervention) and distinguished by color (red, pink, green yellow and peach). Graphs of all the stakeholder networks in this study are shown side by side for visual c omparison in Figure V. 9

PAGE 154

136 Alpha Bravo Charlie Delta Echo Figure V 9 Stakeholder Network Graphs. Local behavioral health center (BH) Early Childhood ( ECC/ E L ) Local school district /BOCE (ED) Local public health agency (PH) Local primary care provider (PC) Local EI service provider (EI) Parent/Family Other nonprofit

PAGE 155

137 Examination of the network graphs highlights patterns in whole network structural characteristics such as the variables of size, density and degree centralization In comparing the graphs in F igure V.9 the differences in network size and density are readily apparent. Stakeholder network Charlie is clearly th e smallest, with the fewest number of organizations, while stakeholder network Bravo is the largest, with the most number of organizations. Comparing their basic structures, it becomes apparent that there is a range of density among the networks and they are all fairly decentralized, as discussed. Aside from the whole network variables, the graphics reveal other findings about organizational positions and connectivity. The local Early Childhood C ouncil (ECC), local public health department or agency (PH ), and local Community Centered Board, the EI service provider contracted by the Colorado Department of Human Services (EI) are often near the center of the networks, with more ties than other organizations ; while the local primary care providers behavioral health center and early childhood preschools are located around the periphery of the networks. This indicates that the ECC, PH and EI organizations likely have higher degree centrality and closeness centrality than the rest of the network membe rs in all five networks In summary, the findings indicate that there are no trends between whole network characteristics and attributes and performance, and the stakeholder networks seem to be arranged similarly. It is thus important to investigate the attributes and characteristics as the organizational level which is the focus of the second research question, to better understand why and how performance differs The structural positions of specific organizations in Figure V.9 provide some clues to t he organizations role s within the

PAGE 156

138 network s Network theorists have also identified organizational level measures that indicate an organization's level of influence in the network, including betweenness centrality, relative connectivity and brokerage, whi ch are investigated within the context of network performance in the next section. Cross Case Analysis Organization Level Role and Influence of Public Health Regardless of overall network integration, there are often one or more actors that are in a position of greater influence in a network. Power and influence in a network are conceptualized and measured in the network literature with relative connectivity, closeness centrality, betweenness centralit y and brokerage measures Table V. 4 summarizes the se key measures for the local public health agency ( PH ) in each network. The measures were also analyzed for two other key network members: the local Early Childhood Councils (ECCs) and the local EI prog rams ( EI s).

PAGE 157

139 Table V 4 Measures of Structural Position and Influence of Network Members Organization Type Degree Centrality Relative Connectivity Closeness Centrality Betweenness Centrality Brokerage, Normalized Alpha PH 10 90% 0 .92 0.03 0.29 ECC 10 100% 0 .92 0.04 0.29 EI 10 75 % 0.92 0.01 0.24 Mean 9 73% 0.87 0.02 0.24 SD 2.13 23% 0.13 0.02 0.14 Bravo PH 16 59% 0 .72 0.04 0.76 ECC 22 100% 0 .87 0.01 0.78 EI 11 60% 0 .63 0.12 0.57 Mean 9.04 38% 0.60 0.02 0.57 SD 5.59 24% 0.10 0.04 0.19 Charlie PH 9 93% 1 0.04 0.44 ECC 8 100% 0 .90 0.01 0.32 EI 8 82% 0 .90 0.13 0.45 Mean 7 76% 0.84 0.05 0.24 SD 1.89 22% 0.13 0.04 0.18 Delta PH 15 84% 0 .89 0.03 0.69 ECC 8 52% 0 .65 0.00 0.03 EI 15 94 % 0 .89 0.01 0.69 Mean 9.78 58% 0.73 0.01 0.29 SD 4.66 26% 0.14 0.02 0.29 Echo PH 7 83% 0 .68 0.00 0.19 ECC 12 83 % 0 .93 0.03 0.56 EI 12 100% 0 .93 0.10 0.63 Mean 7.86 57% 0.74 0.02 0.38 SD 2.88 23% 0.13 0.03 0.18 [Mean and Standard Deviation reflect entire network. PH= Local public health department or agency; ECC= Local Early Childhood Council; EI= Local Community Centered Board, the EI service provider contracted by the Colorado Department of Human Services]

PAGE 158

140 Degree Centrality Degree centrality counts the number of ties each organization has with other organizations in the network. The measure is relative to network size, and thus can o nly be compared within networks To enable comparison, categorical levels were also determined for the variables. A level of Hi gh was attributed to measures above the mean by more than the one standard deviation; Moderately High was attributed to measures above the mean, but within one standard deviation; Moderately Low was attributed to measures below the mean, but within one sta ndard deviation; and Low was attributed to measures below the mean by more than one standard deviation. The results indicate that the PH ECC and EI organizations all had moderate high or high degree centrality within their network, meaning that they had more ties than most organization s in the ir network Relative Connectivity Building upon the degree centrality measure relative connectivity measures both the number and quality of connections between organizations in each network. The measure is based upon: a) b) the level of trust that other organizations have for them and c) how much they value thei r relationships with the other organizations While relative connectivity scores ranged from 18 100% within the stakeholder networks, a striking pattern emerged in the findings There was a he local early childhood 0

PAGE 159

141 Figure V 10 Network Performance by Relative Connectivity of ECCs As indicated by the chart and the polynomial trendline (R 2 = 0.5 3 ), the ECC s had the highest relative connectivity in the higher performing networks; while their relative connectivity was lower in the two lower performing networks. In addition, a pattern emerged indicating almost the opposite findi ng with the local early i ntervention programs which had the highest or among the highest relative connectivity scores in the lower performing networks, as illustrated in Figure V.1 1 R = 0.4878 150 170 190 210 230 250 40 50 60 70 80 90 100 110 Performance Continuum Relative Connectivity, percentage Network Performance by Relative Connectivity of ECCs Alpha Bravo Charlie Delta Echo

PAGE 160

142 Figure V 11 Network Performance by Relative Connectivity of EI s As indicated by the chart and polynomial trend line (R 2 = 0. 59 ), the EI service provider s have lower relative connectivity in the higher performing networks; had moderate relative connectivity in the network in the middle of the performance continuum, and high relative connectivity in the lower performing stakeholder networks. The data do not allow for analysis of a causal relationship or statistically significant correlation; however t he p olymer trend line on both graphs supports the pattern across the five networks in this study. This importance and implication of this finding is discussed in Chapter VI. Closeness and Betweenness Centrality The closeness centrality measure can be used to identify those organizations with the shortest paths to all of the other organizations in the network which provides organizations. The measure is an inverse, with the shortest distance = 1 and the longest distance = 0. R = 0.5775 150 170 190 210 230 250 40 50 60 70 80 90 100 110 Performance Continuum Relative Connectivity, percentage Network Performance by EI Program's Relative Connectivity Alpha Bravo Charlie Delta Echo

PAGE 161

143 The variation among organizations scores and among networks varied so that centrality of its highly connected me mbers. The local public health agency ( PH ) in each network had moderate high or high levels of closeness centrality (shortest distance) in four of stakeholder network s (Alpha, Bravo, Charlie and Delta) In stakeholder network Echo, the lowest performing network, PH had a moderate low level of closeness centrality (longer distance). The early childhood council in each network had high levels of closeness centrality (shortest distance) in all but stakeholder network Alpha, where it had a moderate high leve l, and Delta, where it had a moderate low level (longer distance). The early intervention agency had moderate high levels in the three higher performing networks and high levels (shortest distance) in the lower two performing stakeholder networks. Both p ublic health and early childhood councils had the same or higher closeness centrality than the early intervention agency in the top three performing stakeholder networks, while the local early intervention program had the same or higher closeness centrality in the two lower performing networks. Betweenness c entrality m easures varied tremendously between the local public health agency, early childhood council and early intervention agency across the networks, but at least two of the three organizations had high and moderately high centrality in each network This finding indicates that two of the three key actors in the stakeholder networks were in a position between otherwise unconnected organizations in the networ k. This position has a degree of power and can also become a point of failure for the network. These implications will be discussed in Chapter VI. There was no linear

PAGE 162

144 betweenness centrality of its highly connected members. Brokerage Brokerage is supported in the literature as a good measure for influence in a network. A broker is a structural measurement that indicate s the degree to which an organization spans a structural hole in the network s tructure. There was a strong pattern provider position as a broker, as illustrated in Figure V.1 2 Figure V 12 Network Performance by Brokerage of EI s. As indicated by chart and the polynomial trend line (R 2 = 0.84 ), EI programs were in relatively strong brokerage positions in the lower pe rforming stakeholder networks; were in brokerage position s of more moderate strength the network s in the middle of the performance continuum; and less in a brokerage position in the higher performing networks R = 0.824 150 170 190 210 230 250 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Performance Continuum Brokerage, normalized Network Performance by EI Program's Brokerage Alpha Bravo Charlie Delta Echo

PAGE 163

145 In s ummary, the cross case analysis revealed that ECCs had the highest relative connectivity in the three highest perfo rming stakeholder networks in the study. This means that in each high performing network, the ECCs not only had among the highest number of connections, these connections were also of a high quality: the relationships were with organizations that trusted the ECCs and with organizations that the ECCs valued. The local public health agencies (PH) also showed a fairly similar pattern, with higher relative connectivity in the higher performing networks in this study and lower relative connectivity in the lower performing networks in this study, as summarized in t able V.5. Table V 5 ECC and PH Relative Connectivity Stakeholder Network Relative Connectivity ECC Relative Connectivity PH Alpha 100% 90% Bravo 100% 59% Charlie 100% 93% Delta 52% 84% Echo 83 % 83% The network performance should be taken with some caution given the R squared values of 0.53 and 0.59. However, the trend is worthy of further investigation through content analysis presented later in the chapter Conversely, the cross case analysis revealed a negative association in the EI program lower in high p erforming networks. In fact, in the lowest performi ng stakeholder network, the EI had the hi ghest relative connectivity A similar such pattern was found

PAGE 164

146 regarding the EI s brokerage roles as well, which was higher in lower performing networks and lower in high performing networks. Table V 6 Organization Type Relative Connectivity EI Brokerage, Normalized EI Alpha 75 % 0.24 Bravo 60% 0.57 Charlie 82% 0.45 Delta 94 % 0.69 Echo 100 % 0.63 unexpected, and is explored in the content analysis later in this chapter, and discussed in Chapter VI. Building upon theories of network research detailed in Chapter IV, the attributes of primary care providers are investigated in the following section. Integration of Primary Care Providers The integration of primary care providers is measured by the degree to which they have s trong ties with the other key actors in the system (public health and early intervention) based upon activity type A stronger tie is indicated when s urvey r espondents were to classify the kinds of activities that their organiz with every other organization in the network entailed (in the context of standardized developmental screening, referral and follow up for children in their community)? The respondents could select from the following classifications: None;

PAGE 165

147 Cooperative Activities: such as exchanging information, closing the loop on referrals, attending meetings together, and offering resources to partners; Coordinated Activities: Include cooperative activities in addition to intentional efforts to enh ance each other's capacity for the mutual benefit of programs; or Integrated Activities: In addition to cooperative and coordinated activities, this is the act of using commonalities to create a unified center of knowledge and programming that supports wor k in related content areas A stronger tie is indicated when organization engages in coordinated activities, which would include the implementation of standardized screening in well child checkups and a referral and follow through process with the local e arly intervention agency, and integrated activities. By means of example, Figure V.13 depicts a graph of stakeholder network Alpha, where ties represent coordinated activities. Different from the network graphs that were presented earlier in Figure V.9 the overall network connectedness is not the focus here; rather, it is the ties between primary care providers (represented as brown nodes ) and the local public health agen cy ( PH represented as a yellow node ) and the local early intervention program ( EI represented as a peach node )

PAGE 166

148 Figure V 13 Graph of Stakeholder Network Alpha, Coordinated A ctivities. In Figure V.13, the PH agency has ties with all three primary care providers in the direction of the primary care provider, indicating that it work s to enhance the primary e standardized screening and referral. An arrow represents the direction of the relationship as indicated by the survey respondent. T he direction towards the primary care providers is important, as the ties reflect a relationship that goes beyond simply exchanging information, closing the loop on referrals, attending meetings togeth er and offering resources to partners. The direct ed tie means that the PH agency is intentional about its efforts to enhance the capacity for PH Primary Care 3 CCB Primary Care 2 Primary Care 1 ECC Local behavioral health center (BH) Early Childhood ( ECC/ E L ) Local school district /BOCE (ED) Local public health agency (PH) Local primary care provider (PC) Local EI service provider (EI) Parent/Fa mily Other nonprofit

PAGE 167

149 the mutual benefit of programs. The ECC also has ties with all three primary care provide rs, some ties are directed both ways, but all are directed to wards the primary care providers. The EI in stakeholder network Alpha also has directed ties with two of three primary care providers. Figure V.14 depicts graphs of coordinated ties for the five stakeholder networks for comparison.

PAGE 168

150 Alpha Bravo Charlie Delta Echo Figure V 14 Stakeholder Networks, Coordinated Activities. Local behavioral health center (BH) Early Childhood ( ECC/ E L ) Local school district /BOCE (ED) Local public health agency (PH) Local primary care provider (PC) Local EI service provider (EI) Parent/Family Other nonprofit

PAGE 169

151 Upon comparison of the network graphs illustrating coordinated activities, it is clear that primary care providers and other service providers including the local mental health center and early childhood providers (preschools) are largely on the periphery in al l networks. This is a similarity the stakeholder networks share, regardless of performance. However, a strong pattern emerges in the number and direction of ties between primary care providers and the key actors in the network ( PH ECC and EI ) and th e ov erall network performance. Comparing the ties across stakeholder networks, in the context of performance, a striking pattern becomes clear: the primary care provide rs have more ties that reflect coordinated activities with public health, early interventio n and early childhood councils in the higher performing networks and fewer and fewer ties reflecting such activities in the lower performing stakeholder networks. In stakeholder network Bravo the ECC has directed ties with five of the ten primary care providers. The EI has one directed tie with a primary care practice. In stakeholder network Charlie, PH has a directed tie with the one primary care provider in the network; however neither ECC nor EI have a tie. In stakeholder network Delta, the PH has one directed tie to one of the four primary care providers in the network, and the EI and EEC have no ties with any of the primary care providers. Finally, in stakeholder network Echo, the PH EI nor ECC have a tie with the primary care providers in the network These findings have important implications, given the segmentation of the EI system and the call for primary care provider to undertake standardized screening.

PAGE 170

152 Content Analysis Organizational Perspectives: Systems Building vs. Service Provision The study research design allowed for further investigation into the preliminary findings from the network analysis to learn more about the organizations in the stakeholder networks, indicated by high relative connectivity and high broker age, and see if there is a relationship with performance. The most striking finding was the clear distinction of the Early Childhood Council (ECC) as being a key actor in all high performing networks, with structural position and role and their levels of connectivity and brokerage in the network. These same results emerge for PH agencies, but no t to the sa me extent Further, there were a EI position and role and their levels of connectivity and influence in the network s In order to b etter understand and interpret these findings, content analysis was conducted to illuminate the similarities and differences between these organizations All mission, vision and value statements that were publically available on the Internet were coded an d analyzed in NVivo by first order categories meaning that the statements were clustered into common themes. The same process was then repeated to code the data into s econd order themes to ultimately categorize the organization s into two aggregate dimen sions : those with a systems building focus and those with a n individual service provider focus Table V. 7 presents the results

PAGE 171

153 Table V 7 Classifications of Member Organizations. System Focus Individual Focus of Service Providers Local public h ealth agencies (PH) Local early i ntervention p rogra m (EI ) Local Early Childhood Councils (ECC) Local school district or Board of Cooperative Education (ED) Local pediatrician or primary care provider (PC) Local early learning centers/preschool (EL) Local behavioral health center (BH) Two clear dimensions emerged from the qualitative coding process: a systems approach in which systems building or system level collaboration were explicitly stated as a core mission or vision for the organization and a service provider approach that focused first and foremost on providing quality services at the individual level. O rganizations with a system focus emphasized system change in their mission and vision statements as part of what the organization aimed to achieve: they purposefully worked toward system level goals. By contrast, the organizations with an individual service provider focus aimed to provide high quality services to best meet the needs of their clients/patients. Their vision focused on individual care and service. For example, l ocal public health agencies emphasized achieving their mission through collaboration. One PH long endin g with an through people, prevention and partnerships PH Through partnerships and collaboration [we] will lead our From the language in their mission and vision statements, it w as clear that it established the importance of interorganizational relations in its endeavors to serve their residents Local Early Childhood Councils (ECCs) emphasize their aim to improve the systems that serve young children through system level collaboration. O ne ECC

PAGE 172

154 mission stated that : champions for children and families .[by] serv[ing] as a vehicle that brings together agencies o n a seamless system of care for young children The mission an d vision statements of the ECC emphasize that collaboration is essential that the councils are rooted in the values of community, collaboration and leadership on behalf of children from birth through age eight Th e system focus differed from the individual focus of service providers who a re committed first to their clients and patients providing the highest quality of service high quality care For example, one primary care provider described its Evidenc e Driven, Patient Centered, High Quality, Friendly Health Care. vision and mission statements rarely referred to collaboration and partnership, and when they did, it was framed as a method to better serve the clients For example, teamwork was listed as a core value for many health care providers, but they regard it as teamwork between physician and patient to better serve the patient. The local early intervention programs shared com monalities with both dim ensions ; however, they were coded as having the individual focus because their commitment first and foremost, was serving their clients. Their interest in collaboration and work ing with the community was consistently framed as a method to better serve th The mission and vision and value statements referred to dual goals centered on e nriching of clients and Many statements refere nced being a in partnership with build[ing] reciprocal

PAGE 173

155 relationships with community partners that aim to increase opportunity and high quality services. Sy stems building or systems change was not listed as a service or area of emphasis for any of the local early intervention providers. The dimensions, and subsequent classification, of system focused organizations and service provider that emerged from the content analysis was considered in the context of the network analysis findings described previously. Given the varied perspectives and priorities of the organizations involved in each stakeholder network, the ap plication revealed to the a structural position and influential role of the system builder s in the networks The higher performing networks had a t least one system focused organization that was well connected and trusted, in the powerful and influential position to broker relationships and facilitate flow within the network. By contrast, the lower performing networks had a service provider in this position. The findings do not indicate that any one organization is any more capable than other; it is the perspective and emphasis that they bring to the network in that position The findings suggest that the perspective of a key organization in the stakeholder network, measured by high rela tive connectivity, is related to performance. Conclusion In summary, the findings indicated that there were no relationships in whole network characteristics relative to network performance. It was also shown that trust, which might be necessary to network performance, as indicated in the literature, was not sufficient for performance.

PAGE 174

156 The stakeholder networks tended to be organized around the local Early Childhood Councils (ECCs), local p ublic health agency (PHs) and local early intervention program (EIs). Two of the three organizations served as connectors within each stakeholder network. While both public health and early intervention are federally legislated agencies in the EI system and were thus expected to be central actors in the networks, the Early Childhood Councils central role in the networks was an unexpected finding. The positive relationship between the ECC relative connectivity and network performance was indicated in a clear pattern in the cross case analysis. Unexpectedly, networks with EI programs in a broker position had poorer performance. These contrasting patterns were further investigated through content analysis. T he distinction between the types of organizat ional perspective emerged from analysis of organization mission, vision and value statement, which helped to illuminate the importance of an to network performance The f inding s indicate d a positive relationship between network struct ures with a systems focused organization in a key position and overall network performance ; while findings indicated a negative relationship between network structures with a service provider in a key position and overall network performance Finally, stro ng relationships between the three central actors (ECC, PH and EI) and primary care providers, as measured by coordinated activities to efforts to enhance mutual benefit, was found to be positively related to network performance.

PAGE 175

157 CHAPTER VI VI. DISCUSSION Linking Network Attributes and Characteristics to Performance This chapter discusses the research findings presented in the preceding chapter. A brief summary of the study is offered; the findings are then interpreted with regard to the literature and discussed in the context of the study research questions Conclusions for health and public health systems are then offered, as are recommendations for the application of study findings and future researc h Summary of Investigation The reliance upon interorganizational relationships in health and public health systems is extensive; however, there is a shortage of evidence that focuses on stakeholder network performance. Responding to this gap and the call for a systems approach, the ask ed whether community based stakeholder networks operating within the same context exhibit commonalities in network level attributes and characteristics relative to their performance. The second question investigated organizational level variables, asking if the relative connectivity and structural positions of member organizations affect s takeholder network performance In a mixed method, cross case research design, the characteristics a nd attributes of five stakeholder networks were analyzed and compared Whole network, dyad and organizational measures were examined through social network analysis (SNA) and content analysis and analyzed with performance measures based on stakeholder in terviews, self assessments and local early intervention assessments. Focusing on the dynamic interactions of public health agencies with public and private organizations that

PAGE 176

158 affect health, the five stakeholder groups were goal directed networks, working to ensure standardized developmental screening, referral and follow up for children in their Colorado communities The stakeholder networks were mobilized to overcome the segmented system of early intervention services. The study found that there were n o patterns in the cross case analysis of network level variables relative to performance ; while linear patterns were found in organizational level variables relative to performance T he findings showed that high performing netw orks had system building organization s in position s with high relative connectivity while the lower performing network s had a service provider in those position s Primary care providers also had strong ties with the key actors in h igh performing stakeholder networks and these t ies gradually decreased in number when compared with networks lower along the performance continuum. Results also showed that trust is necessary, but not a s ufficient condition of network performance. Exploring Research Question One This study first asked: Do community based stakeholder net works operating within the same c ontext exhibit commonalities in network level attributes and characteristics relative to their performance? This research falls into an emerging area of network research that focuses on the relationship between network structures at the whole network level, such as density or network degree centralization, and overall network performance, again measured at the w hole network level (see Provan & Milward 1995, Rosenheck et al 1998; Provan et al., 2007).

PAGE 177

159 Network Structure While a model of network effectiveness (Provan & Milward, 1995) exists in the literature that correlates whole network measures of integration with overall network effectiveness, subsequent research that focused on the validation of this work has found contradictory findings ( Rosenheck et al. 1998). Further, l ow density/high centrality networks are typical of innovation networks, but the relat ionships between network integration and performance have not be empirically tested (Dhanaraj & Parhke, 2006). This area of organizational network research is still fragmented and under developed, and research available to inform practice & Lemaire, 2012, p. 653.) Thus, without a clear model in the literature, the first research question sought to explore structural integration. Specifically, the research contributes to the literature by analyzing the relationships between network integration and performance. The research design enabled a unique comparison of five locally formed stakeholder networks operating in their loca l syste m of early intervention that all shared the same goals and fairly stable levels of resource to explore commonalities and patterns. Network integration was conceptualized by : network density, which provides insight into the overall basic connectedness of an organization ; and the degree of centralization, which illuminates to the degree to which the network is structured around certain organizations relative to others Utilizing the most different cross case design, the findings included considerable range in both measurements, although none of the networks were highly centralized

PAGE 178

160 The findings indicated that community based stakeholder networks operating within the same context exhibit do not exhibit commonalities in network level attributes and characte ristics relative to their performance. T he findings revealed no patterns in network demographics or whole network variables relative to performance. Further, there was no relationship between level of integration concep tualized with whole ne twork measurements of density and degree centralization and its performance The finding s are counter to Proven and Mil effectiveness; however, that model has not yet been validated by other researchers. Further, Proven and Milward based their model upon research focused on a service delivery network, which is a different type than the stakeholder ne tworks in this investigation The absence of a pattern is signif icant for health and public health systems research, where policy makers and public health leaders are looking for guidance around the most effective and efficient way to leverage limited resources through organizational network s. Although a topic of great interest in network literature there is no clear answe r about what type of organizational network structure is most effective T he findings in this study suggest that whole networ k integration does not drive performance. While the highest performing network had a high density and low degree of centralization, another high performing stakeholder network had a low density and comparatively high degree of centralization. purpose and formation, it is surprising that they varied so much in whole net work integration, and that there was no pattern in relation to their performance P erhaps the of network integration has yet to be formulated in

PAGE 179

161 organizational network ution. Consideration of just the structure and integration of organizational networks may be too simplistic of a perspective to account for the complexities within the dynamic social structures of organizational networks. The study findings indicate that organizational level attributes and characteristics may be what really matters to network performance. Network Trust Network and collaboration research literature emphasize the importan ce of trust for effective interorganizational relationships and networks ( Uzzi, 1996 1997; Bryson et al., 2006; Thomson et al., 2008b) Further, trust is a common theme in virtually all definitions of interorganizational network s (Provan et al ., 2007 p. 643 ), with empirical evidence that overall trust is an integral condition for network performance (Uzzi, 1 996; Provan et al., 2003 ; Zaheer et al ., 1998; Huxham & Vangen, 2005; Ring & Van de Ven, 1994; Bryson et al., 2006; Thomson et al. 2008 b ). Based upon these findings, one would expect to see a pattern of higher trust in the higher performing stakeholder networks, and a lower trust in the lower p erforming stakeholder networks. Trust and performance were analyzed for each stakeholder network at the whole network level to identify if such a pattern exist ed in the cro ss case analysis. Study results showed that all five stakeholder networks in this investigation had fairly high levels of trust, ran ging from 67.40 78.70 percent ( where 100 percent trust would mean that all members of the network trust ever y me mber at the highest level ) (Varda, 2012). Further, there was no pattern in performance relative to trust: trust levels were high, regardless o f performance

PAGE 180

162 that trust is an int egral conditi on for performance ; however, the findings reveal that trust is not a sufficient condition for network performance. The most different cross case design of this study enabled disconfirming research finding stakeholder networks with high trust levels but lower performance Since trust levels were reported high in all the stakeholder networks, including those with poorer performance, the findings suggest that either trust is not a key driver of performance, or it may be over estimated by survey respondent s This result has significant application in health and public health system s where network leaders and members often spend considerable time building trust within their organizational network s or collaboratives. It is important for practitioners to un derstand that those efforts, while vitally important, will not be enough to develop and manage high performing stakeholder networks. In summary, the findings in response to research question one indicate: 1) no whole network relationship to performance; and 2) trust levels were counter to expectation and are an insufficient condition for network performance. These findings suggest that something else explains network performance. Deeper analysis of the dyads between and among organizations within the ne tworks i s the focus of the second research question Exploring Research Question Two Do the relative conne ctivity and structural position of member organizations affect stakeholder network performance ? This research focuses on the im pact of an s relational and structural attributes and characteristics, measured at the individual member and dyad level, as opposed to the whole network level. Seeking to better understand correlations between

PAGE 181

163 organizational level variables and overall network outcomes this second question focus es the investigation on th e impact of the structural positions of certain types of organizations, how they are connected, and the roles they play in the network Organization Positions and Roles As detailed in Chapters I and II, the fundamental charge of public health is 1988, p. 140). Further, t he Social Security Act, Title V Maternal and Child Health Pro gram legislates that public health agencies implement family centered, community based, systems of coordinated care for children with special healthcare needs (MCHB, 2012a). Since local public health agencies offer few direct EI services, they must coordi nate, connect and collaborate with primary care, human services and education to ensure effective identification, screening and delivery of services for children with developmen tal delays This is especially important given that health and pub lic health l iterature identify challenges with the segmented system of early intervention. Thus it would be expected that public health agencies ( PH s) would be in a central and highly connected position in high performing networks, and that they would take on a broke r ing role in the sta keholder networks, working to connect organizations tha t make up the disconnected part s of the system. Counter to expectation, t he study findings revealed that public health agencies had a fairly central role in most networks; however, they were not necessarily the most connected or in the strong est broker age position s A striking finding of the investigation was the emergence of the early childhood councils (ECCs) in these roles. As discussed in Chapter V, the ECC s had the highest

PAGE 182

164 relative connectivity in the three highest performing stakeholder networks in the study. This means that in each high performing network, the ECCs not only had among the highest number of connections, these connections were also of a high quality : t he rel ationships were with organizations that trusted the ECCs and with organizations that th e ECC s valued. The positive association between their high relative connectivity and the network performance thus integral role in the stakeholder networks. This finding builds upon the organizational network literature, where research correlate s Hanneman & Riddle, 2005 ). In the higher performing stakeholder networks, the ECCs were all well connected and us ed these connections to broker new relationships between previously unconnected parts of the networks. In doing so, they buil t influence within the network (Hanneman & Riddle, 2005) While no found in the organizational network literature, in the context of these networks, it appears networks to achiev e higher performance. With such high centrality and trust for the ECCs, they may be the ideal broker organization, with the ability to connect organizations and to pass along important information. cessibility, capacity and (Colorado Department of Education [CDE], 2013, para.1) This systems building focus may enable them to more effectively and efficiently bring tog ether and work with partners Systems building efforts are rarely funded or measured in the public sector and

PAGE 183

165 organizations often balanc e so many directives that they need to prioritize funded and measured responsibilities first. ECCs are in the unique position of being able to focus entirely on systems level efforts, which is aligned with their core mission, funding and evaluation. This may contribute to their success as an influential connector in the high performing stakeholder networks. T he fact tha t ECCs were in these roles more than public health agencies highlights two opportunities for health and public health leaders : a) The findings provide an opportunity to further consider the role of ECCs in local early intervention systems since they were an integral actor in these stakeholder networks in a system in which they may not always be considered a primary stakeholder by policy makers b) There is an opportunity for public health agencies to better leverage their multiple community connections and syst ems perspective to further improve the EI system in their communit ies by assuming a more active connector and/or broker role Conversely the local early intervention program ( EI ) was in an almost inverse position to the ECCs relative to network perform ance. The cross case analysis revealed performing networks and lower in high performing networks. In fact, i n the lowest performing stakeholder network the EI had the highest relative connectivity, meaning that they had the highest number of quality connections with other organizations in the network.

PAGE 184

166 A similar pattern was found regarding the EI s brokerage role s as well, which was highe r in lower performing netw orks and lower in high performing networks. Kilduff and Tsai (2003) found that the role of a broker is only advisable for organizations that the EI brokerage role i s surprising and counter to expectation given the legitimacy that the EI service provider would bring to a network focused on EI. Fernandez and Gould (1994) found that in order to be an effective broker in networks in the policy arena, actors needed to be perceived as relatively impartial to the policy initiatives Although the network s in this study are goal directed stakeholder network rather than policy network s a similar perception may be influencing EI provider s ability to be successful brokers. to improve the segmented system for the benefit of children Both organizati ons are working toward the same goal, but the primary focus of their organization may shape the way their message and efforts are received by the other organizations in the networks. The stark contrast of: the positive relationship between network structures with ECC in a key position and network performance; and the negative relationship between network structures with EI in a key position and network performance, indicated that something abou t the differences between those two organizations was diving network performance Content analysis identified the primary distinction between the organizations from their perspective. Both the ECCs and PH s have a system building focus as a core part of t heir mission s and organizational values. The EI s on the other hand while committed to community engagement and partnerships to better serve their

PAGE 185

167 clients, first and foremost focus on providing quality service s on an individual basis All the other stak eholder network members were services providers, also focused first on serving their clients. This distinction is significant in light of the cross case network analysis findings : all high performing stakeholder networks had an organization with a systems building perspective in an influential and central role as evidenced by high relative connectivity and brokerage measures ; whereas the lowe r performing stakeholder networks had a service provider in an influential and central position I n fact, it was a n inverse pattern Contributions to the Literature An important conclusion of this study is that one cannot consider the whole network without examining the structure of the component parts. Th ese findings provide important insight in to organizational net work literature, as empirical research that & Provan, 2009, p .2; Provan et al. 2007 ; Provan & Kenis, 2008 ). This area of study is fundamentally about understanding the impact of network attributes, characteristics and conditions on network outcomes, and although the research is quite rich in describing various positions in a network, little attention has been paid to the type of organization in such a role, and its impact on network performance. This study provides insight not just in relation to various member organization s impact on the network, but to whether the organization maintains a systems building or service provision emphasis in the network Another important finding in this study is th e strong distinction between systems focused organizations and service provider organizations as key connector s and broker s in the networks and their impact on network performance. This is an area that we still do

PAGE 186

168 not fully und erstand and requires further research. It may be that the service focus impacts the perception of the other organizations in the network, and aff ects willingness to stay involved and committed Recall that EI ne eds the rest of the system to effectively function so that children with developmental delays are identified early and referred to them for services. EI providers may be perceived as telling everyone what needs to be done so that they can deliver services, whereas the systems focused organization may be perceived as wanting improve the systems for A third consideration is that the service providers simply do not have the skills or capacity to serve as effective network connectors or brokers. Service providers, by definition, are focused first and foremost o n providing quality services to their clients and/or patients. They are often compensated only through direct service provision and may not have the time, resources, expertise, skills, or even interest, in playing an important connecting role in the networ ks. This is another under investigated area in the organizational network literature (Conklin, Lusk, Harris, & Stolee, 2013). Conklin and colleagues recently set out to define and describe the role of brokers in a knowledge exchange network and found qui te simply that "you need people with skills to enable groups to come together" (Conklin et al., 2013, p.5). Their findings indicated that most knowledge brokers described their primary role as connecting actors with other actors, information, ideas, and e ducational resources; and the ir secondary role as coordinating actors Other network actors identified brokers as important in promoting mutual understanding among network actors (Conklin et al., 2013). Along the same lines, Miles and Snow (1986) found th

PAGE 187

169 coordinates and facilitates organizations within networks, but does not command. By their very nature, the systems focused organizations often take on a coordination and facilitation role in their day to day systems building work Further investigation i s required to understand whether the systems focused organizations are thus better equipped to take on such roles in networks than the service providers. The goal congruence of a network broker an d the overall network should also be considered in light of th e study finding s The stakeholder networks in this study aim to ensure standardized developmental screening, referral and follow up for children in their community. Thus, the network s are work ing to improve the system that leads up to the provision of services along with the connection to the EI service provider As identified in the content analysis, the EI efforts are focused on providing quality EI services to the children who are referred to them. Although EI provide rs are an integral stakeholder in the EI system, their goals primarily focus on a different part of the system (the service delivery) than the networ ks goals (the identification, referral and follow though) Lundin (2007) found that goal congruenc e affects cooperation among organizations and specifically, that organizations must have both trust and goal congruence for effective joint action. Althou high trust scores demonstrate that the organizations are perceived as trustworthy by other network members, the misalignment of their goals in relation to network goals may be inhibiting their ability to be an effective connector and bro ker. ECCs, on the other hand, aim to improve the systems that serve young children through system level collaboration. The align s perfectly with the ECCs Also a highly trusted member of each network, the ECCs have

PAGE 188

170 both trust and goal congruence, which may enable them successful connect and broker key relationships in the stakeholder networks to ultimately enable the networks to perform better. The affect of goal congruence on network roles is not yet well understood in the literature. Finally the notion of accountability of the network members is also worthy of further consideration in light of the findings. The EI providers in Colorado are accountable to the Colorado Depart ment of Human Services (CDHS), who are accountable to Congress, for the delivery of high quality and timely EI services to children with developmental delays. Thus, i t may not be appropriate to expect them to spend their efforts and resources to be effective connectors or brokers in stakeholder network s Although the networks aim to improve the EI system through enhanced screening, referral and follow up for children the legislated requirements of the EI providers focuses only on the provision of services ( to th ose referred) Thus, involvement in the se networks is not strictly part of their legislative charge. Of course, the networks aim to overcome the segmentation in the system and improve the outcomes for children with developmental delays and so their involvement is neces sary for the networks to succeed but their participation as a connector or broker may be stifled by the fact that they are not accountable to these networks, but rather to CDHS and Cong ress. On the other hand, ECCs and PH agencies are accountable for systems level change, and thus the performance of the network is congruent with their accountability. The findings also have significant application for health and public health systems wh ere a key focus is on getting the right organizations to come to the table and be a part of an interorganizational effort The findings illustrate that it is not simply a

PAGE 189

171 matter of having th e right organization s engaged in the network they must also be in the right positions The networks in the study, w hich are working to ensure standardized developmental screening, referral and follow up for children in their communities are goal directed stakeholder networks with an aim to improv e the local segmentation a n d implementation of the EI system. For such networks, the study results sho w that having an organization with a systems building perspective in key and influential position is crucial for network performance. These findings hel p to define what a systems focused organizations is, in very tangible way. Strength of Ties Given the segmentation in the EI system and the call for integration of primary care providers, one would expect to see primary care providers well connected in th e high performing networks. Network research tells us that being well connected does not mean that an organization needs to have ties with all members of a network. It is important to be strategic in the number, type and direction of ties, but also in th e strength of ties. Karckahrdt ( 1992 2010 ) argues the benefits of strong ties, while Granovetter (1973 1992 ) found that there are benefits in weak ties Further, r epresentatives of organizations have a limited amount of time and energy available for maintaining ties with other organizations especially strong ties which typically require more of an investment of time and energy. It is generally much easier to maint ain a high number of weak ties (Provan & Lemaire, 2012) Building upon the literature, it was expected that primary care providers in high performing networks would have stronger relationships with public health and early intervention than with other orga nizations in the network. Given the t would be important that the relationships not

PAGE 190

172 simply indicate an awareness of the organizations or frequency of contact; rather the ties should represent coordinated activities to ensure system effectiveness. Thus a strong relationship was conceptualized as a relationship that included cooperative activities, in addition to intentional efforts to enhance each other's capacity for the mutual benefit of programs. Comparing the rela tionships to and from primary care providers with other organizations in each stakeholder network a striking pattern became clear: the findings revealed a pattern in the number of strong relationships between primary care providers and public health, earl y childhood councils and early intervention. Specifically, the PH s, ECCs, and to some degree, EI s, had more relationships that reflected coordinated activities with primary care providers in the higher performing networks and fewer relationships that reflected such activities in the lower performing stakeholder networks. The positive association of the primary care providers strong ties with key actors and the helps to illuminate which relationships are worthy of investment to f oster and maintain as strong relationships and where weak relationships are sufficient. The direction of these strong ties towards the primary care providers is also important, indicating that the ECCs, PH s and EI s were working to enhance the primary ca re p standardized screening and referral and follow through The findings care providers within an early intervention stakeholder network impacts the network contribution to public health strategies at the local levels. The finding s build upon the network literature to add to the understanding of the role of strong and weak ties in networks. Uzzi (1997) found that s trong ties enable

PAGE 191

173 increased proble m solving capacity ; Hansen (1999) found that they can provide richer and more effective transfer of info rmation; while Karckhardt (20 10 ) found that strong ties can be trusted sources of influence. However, organizations are limited in the number of strong ties they can maintain, and strategic decisions about with whom to have strong ties is important. Research has brought tremendous insight into both strong and weak ties, but a greater understandin g of an effective mix of strong and weak ties is also critical. The findings show that key relationships mattered in this study: strong relationships between disconnected parts of the EI system were important for network performance, and ultimately EI sys tem performance. There was no pattern relative to performance for sys tem context are considered. These findings have significant implications for health and public health systems, given the segmentation of the EI system and the call for primary care provider s to undertake standardized screening. The study results indicate t hat st rong ties with ECCs, PHs and EI s may help primary care providers to implement such screening tools to better identify children with developmental disabilities, thereby helping the network to achieve its goals and strengthening the local EI system. Wh ile the health literature indicates that up to half of all children with developmental delays are not identified until they enter school (Mackrides & Ryherd, 2011), there is a consensus in the field that pediatric physicians have both the opportunity and e xpertise to identify children in need of EI services and are thus the ideal candidates for helping to improve the segmented system (AAP 2001). These findings are also in line with the recent and significant attention

PAGE 192

174 focused on medical homes as a mechan ism to help ensure quality in health care for children with disabilities, developmental delay or other special health care needs. The medical home model requires that children receive family centered care from a physician or nurse who can also provide read y access to referrals and care coordination when needed (Strickland et al., 2011). Many pediatricians and family medicine physicians do not regularly use standardized screening tools, even though there is evidence that a screening tool will more effectively and accurately identify children who may have a develop mental delay ( HHS, 2000; Rydz et al., 2005). Thus t he directiona lity of these ties is also paramount given the need to integrate primary care physicians into the EI system for the important role of identification, screening and referral. The findings show that the ECC s PH s, and to some degree, the EI s were intention al about their efforts to enhance the primary care The findings reinforce the importance of thinking strategically not only about the types of relationships, but also about the purpose and cost and be nefits of the relationships in a network aimed at improving health and public health outcomes. Conclusions for Early Intervention Health and Public Health Systems Interorganizational relationships are inherent in the health and public health systems where public, private and nonprofit organizations work together to ensure that America becomes "a society in which all people live long, healthy lives" ( HHS, 2012, Message from the Secretary, para. 1) It is thus so important that policy leaders and practition ers are armed with the knowledge to prompt strategic and thoughtful discussion to foster stronger inclusion, involvement and commitment where it most makes sense

PAGE 193

175 and will have the greatest impact. This study aimed to provide insight into stakeholder netw ork performance in the context of early intervention, which crosses the health, public health systems, along with education and human services systems. More specifically, t he attributes and char acteristics relative to its performance, adding to a much needed knowledge base to guide the use, formation and management of networks. Reframing the study findings to that end, these three conclusions are offered for consideration. Conclusion 1 : Perspecti ves of Key Actors Matters The diverse group of public and private organizations and other stakeholders detailed in this dissertation work together in multiple settings with overlapping health related goals and missions, ultimately trying to contribute to the health of society (IOM, 2012a). For goal based networks, it is integral not only to align the network membership with the goal s the network aims to achieve, but to align the perspectives of key actors with the network goals. Specifically for networks aiming to improve health and public health systems, a systems level focus of key actors may be critical for network performance In this study of goal directed, community based stakeholder networks; key actors in the network that had a systems focus were positively associated with network performance, while key actors that were service providers were not as well suited for the se ke y positions or role and were negatively associated with network performance Conclusion 2 : Integration of Primary Care c an Strengthen Outcomes As with any research of interorganizational relationships, it is important to consider the context of the rel ationships under investigati on. As detailed in Chapter II, stakeholder networks were mobilized to overcome the segmented system of identification

PAGE 194

176 and delivery for early intervention (EI) services by, in part, helping to bridge a gap between primary care with public health, education, and human services. Recall that state and local agencies from public health, human services and education all must wo rk together to implement federal policy to ensure a system of delivery for early intervention services However, these agencies do not typically i dentify children for EI services A systems approach highlighted the fact that the identification and referr al to EI was a gap in the system because it was not addressed by the legislation Further, without well defined inter agency and inter system processes, there are opportunities for children to be lost between systems. For these reasons, primary care prov iders have being called upon to fill this gap, and thus it is vital that they are integrated in the stakeholder networks. A considerable challenge they face is demonstrated in network and collaboration research: that it is unfeasible and unsustainable to maintain numerous strong ties (Krackhardt 2010 ). This would especially be the case for physicians, as their primary focus is on providing individual care. This is fu rther supported by Varda et al. only difficult for publ ic health leaders to maintain interactions with collaborating organizatio ns, but likely unsustainable. Furthermore, PCs are service, not system, oriented. Taken together, it would be unfeasible to expect that primary care providers could build and sustai n many strong relationships in these stakeholder networks However, it is critical that they are engaged to be able to bridge the segmentation that exists in EI and enable the network to leverage its collective strengths and achieve its goals.

PAGE 195

177 Strategicall y focusing on how to connect primary care providers, and which relationships are worth the investment of time and effort to build and maintain is critical, so that relationship budgeting can make the most out of the primary care providers involvement. Th is study revealed that when PH s and ECCs, and to some degree, EI s network performed well, and the entire system benefited. Conclusion 3 : Bigger Is Not Necessarily Better Rather than simply buil ding a large network, Burt (1995 ) has argued that the pattern of ties is more important than the size of one's network. The s tudy findings its size T his is significant in health and public health systems where organizational network s and efforts are measured in part by simply counting members As briefly discussed in Chapter I, the National Public Health Performance Standards Program (NPHP SP) assesses PH s by the degree to which they are: Identifying potential stakeholders who contribute to or benefit from public health, and increase their awareness of the value of public health ; Building coalitions to draw upon the full range of potential human and material resources to improve community health ; and Convening and facilitating partnerships among groups and associations (including those not typically considered to be health related) in understanding defined health improvement projects, including preventive, screening, rehab 2012, NPHPSP, Local Instruments: Local Public Health System Performance Assessment).

PAGE 196

178 These assessments provide no guidance or incentives for being strategic about the relationships between and among organizations in a stakeholder network. Instead of expecting the PH to connect and interact with every organization that might possibly supp ort overall population health in some way, organizational network research can be used to distinguish and facilitat e connections as detailed in this chapter so that strong ties are established and fostered where they can have greatest impact, and weak ti es are maintained so they can provide access to new information For example, PH agencies may be most efficient serving as a broker in systems building network efforts, but may be most efficient as a peripheral member of a service delivery network. Organi zational networks in health and public health hold the promise of enabling more efficient and effective use of resources through strategic leveraging, economies of scale and collective action (Bryson et al ., 2006) ; however, further research is needed in th is area to guide strategies to enable networks to be as effective and efficient as possible Future Research Needs Interorganizational relationships and networks are a large part of the public health 2012, S72). The more evidence that can be produce d about how the characteristics and attributes of partnerships affect their performance, the more public health strategies and services can be organized, financed and delivered to leverage the strengths of these interorganizational relationships while minimizing their weaknesses. This investigation focus ed on applied research with the aim to contribute insight to early intervention, health and public health leaders and practitioner s by investigating how different attributes and characteristics o f

PAGE 197

179 the whole network and the relationships between organizations within the network are linked to network performance. Specifically, the research contributes to the literature by assessing t he relationship between different characteristics and attributes of stakeholder networks and network performance with the conclusion that one cannot consider the whole network without examining the structure of its component parts. The strength of this st health and public health systems comes from the mixed method cross case analysis of five similar, but locally formed goal directed stakeholder networks. The findings revealed no whole network relationship to performance, while they indic ated a positive relationship between network structures with a systems focused organization in a key position and overall network performance; and a negative relationship between network structures with a service provider in a key position and overall netw ork performance. an integral or influential role for improved network performance There is much more to learn about this notion, and this would be a fasci nating n ew research direction. This study could be replicated with a greater number of cases or different types of networks to investigate whether the patterns hold. One could use an entirely different research design to investigate where there are one or more ty pe s of organization s that can most effectively lead organizational networ ks and why The distinction between service providers and organizations with a systems focus emerged from the network and content analysis The need for systems level change in public health is reinforced in the literature, yet s ystem building efforts are rarely funded or measured in the public sector P ublic management literature demonstrates that

PAGE 198

180 managers prioritize those tasks that are measured and that have funding attached. A new approach to better understanding the implications of systems level funding could investigate the funding trail of federal, state and private sector funds dedicated toward change at the system level or focus on the accountably of organizations with in systems building efforts In this study, ECCs are in the unique position of being able to focus entirely on systems level efforts, which is aligned with their core mission, funding and evaluation. This may contribute to their success as key actor s in the high performing stakeholder networks The fact that EI programs are funded for the individual services provided, as in primary care, may have contributed to their lack of success as key actor s in the stakeholder network. Furthermore, these findings raise new questions about why EI providers were not effective connectors or brokers in the stakeholder networks. Was it the way they were perceived by the network members because of their service delivery focus ; the misalignment of their goals with the ne l ; or was something else altogether ? This is an area that is not yet fully understood, but the study findings certainly show that something is at play here and requires further research The study findings also reveal that there may be a competency that network brokers require in order to be effective. That begs the question: is a certain skill set needed to broker or govern organizational networks? This area is underdeveloped in the network literature and would be a fascinating are a of future research Finally, this study found that trust, while consistency linked with performance in the collaboration and network literature, was not a sufficient condition for network performance. This finding bring s on many questions about what else is needed, in addition to trust, to be sufficient for performance. Building upon that, would such a

PAGE 199

181 condition be strong enough to replace, or supersede the importance of trust? One could also investigate if there was a point of diminishing return on trust ? Also, h ow would a lack of trust affect these stakeholder groups? Is trust between organizations difficu lt for the employee s or representatives of an organization to critically assess? Is it over estimated in organizational network research? These kinds of questions have yet to be answered, but could inform the way organizations work together to most effectively and efficiently achieve outcomes that are not possible by working alone.

PAGE 200

182 REFERENCES Agency for Health Care Research and Quality [ AHRQ ]. (2008, October ) Portfolios of r esearch Retrieved from http://www.ahrq.gov/fund/portfolio. htm Agranoff, R. ( 1986 ) Intergovernmental management: Human service problem solving in six metropolitan areas Albany: State University of New York Press. Agranoff, R. & McGuire, M. ( 1998 ) Multinetwork management: collaboration and the hollow state in local economics. Journal of Public Administration Research and Theory, 8 (1) 67 92. Agranoff, R. & McGuire, M. ( 2001 ) Big questions in public network management research. Journal of Public Administration Research and Theory, 11 (3) 295 326. Agranoff, R. & McGuire, M. ( 2003 ) Collaborative public management: New strategies for local governments Washington, DC: G eorgetown University Press. Ahuja, G. (2000). The duality of collaboration: Inducements and opportunities in the formation of interfirm linkages. Strategic Management Journal, 21 (3), 317 343. Ahuja, G., Soda, G., & Zaheer, A. (2012). The genesis and dynamics of organizational networks. Organization Science 23 (2), 434 448 Alter, C., & Hage, J. ( 1993 ) Organizations working together. London: Sage. American Academy of Pediatrics [AAP] Committee on Children with Disabilities. (2001). Developmental su rveillance and screening of infants and young children. Pediatrics 108 (1), 192 6. American Academy of Pediatrics [AAP]. (2013). AAP f acts. Retrieved from http://www.aap.org/en us/about the aap/aap facts/Pages/AAP Facts.aspx American Academy of Family Physicians [AAFP]. (2013). About us. Retrieved from http://www.aafp.org/online/en/home/aboutus.html?navid=about+us Assuring Better Child health and Development [ABCD]. (2011). ABCD model community guide Denver, CO. Assuring Better Child health and Development [ABCD]. ( 2012 a ) Assuring Better Child h ealth Development (ABCD): About us Retrieved from http://coloradoabcd.org/ Assuring Better Child health and Development [ABCD]. (2012 b ). Assuring Better Child health Development (ABCD): Health care providers. Retrieved from http://www.coloradoabcd.org/healthcare/index.html

PAGE 201

183 Assuring Better Child health and Development [ABCD]. (2012 c ). Assuring Better Child health Development (ABCD): Child care providers. Retrieved from http://www.coloradoabcd.org/pro viders/index.html Barabsi, A. L. (2002). Linked: The new science of networks. New York: Perseus Books. Barnett, W. S. (1995). Long term effects of early childhood programs on cognitive and school outcomes. The F uture of C hildren 5 (3) 25 50. Bennett, J. (2008a). Early childhood services in the OECD countries: a review of the literature and of current policy in the early childhood field, Innocenti Working Paper No. 2008 01. Florence: UNICEF. Retrieved from http://www.unicef irc.org/publications/pdf/iwp _2008_01_final.pdf Bennett J. (2008b). Early childhood education and care systems in the OECD countries: The issue of tradition and governance. In Tremblay, R.E., Barr, R.G., Peters, R., Boivin, M., (Eds.), Encyclopedia on Early Childhood Development Mont real (pp.1 5 ) Centre of Excellence for Early Childhood Development. Retrieved from http://www.child encyclopedia.com/documents/BennettANGxp.pdf Berg, B.L. (2004). Qualitative research methods for the social sciences (5th ed.). Boston, MA: Pearson Education. Berry, F. S., Brower, R. S., Sang Ok, C., Goa, W. X., HeeSoun, J., Myungjung, K., et al. (2004). Three traditions of network research: What the public management research agenda can learn from other research communities. Public Administration R eview, 64 (5): 539 552. Berry, C., Krutz, G. S., Langner, B. E., & Budetti, P ( 2008 ) Jump Starting Collaboration: The ABCD Initiative and the Provision of Child Development Services through Medicaid and Collaborators. Public Administration Review 68 (3), 480 490. Bethell, C., Reuland, C. H. P., Halfon, N., & Schor, E. L. (2004). Measuring the quality of preventive and developmental services for young children: national estimates Pediatrics 113 (S5), 1973 1983. Bingham, C. (2003). An assessment of health service needs for the African American c ommunity in Lexington, KY. Retrieved from http://ahec.med.uky.edu/ Boards of Cooperative Educational Services [BOCES] (2012). Who Are We? Retrieved from http://www.colora doboces.org/ Bogason, P. & Zolner, M. ( 2007 ) Methods in democratic network governance New York, NY: Palgrave Macmillan.

PAGE 202

184 Borgatti, S. P. (2012). University of Kentucky's LINKS Center workshop for s ocial n etwork analysis. Presentation June 3 7, 2012. Lexington, KY Borgatti, S.P., Everett, M.G. & Freeman, L.C. ( 2002 ) UCINET for Windows: Software for Social Network Analysis [computer software] Harvard, MA: Analytic Technologies. Borgatti, S. P., Jones, C., & Everett, M. G. (1998). Network measures of social capital. Connections 21 (2), 27 36. Brass, D. J., Galaskiewicz, J., Greve, H. R., & Tsai,W. ( 2004 ) Taking stock of networks and organizations: A multilevel perspective. Academy of Management Journal 47: 795 817. Bright Tots. (2012). Early Intervention. Retrieved from http://www.brighttots.com/early_intervention.html Bryson, J. M. Crosby, B. C. & Stone, M. M. ( 2006 ) The Design and Implementation of Cross Sector Collaborations: Propositions from the Literature. Public Administration Review 66 ( S1) 44 55. Burt, R. S. ( 1995 ) Structural holes: The social structure of competition Boston, MA: Harvard University Press. Burt, R. S. (1997). The contingent value of social capital. Administrative Science Quarterly 42 (2), 33 9 365. Burt, R. S. (2000). The network structure of social capital. Research in Organizational Behavior 22 345 423. Burt, R. S. (2001). Structural holes versus network closure as social capital. In Lin, N., Cook, K. & Burt, R.S. (Eds.) Social Capital: Theory and Research (pp. 31 56). New York, NY: Aldine de Gruyter Butterfoss, F. D., Goodman, R. M., & Wandersman, A ( 1993 ) Community Coalitions for Prevention and Health Promotion. Health Education Research 8 (3) 315 330. Catalog of Federal Domestic Assistance [CFDA]. (2013). Maternal and Child Health services block grant to the States: MCH block grants. Available from https://www.cfda.gov/index?s=program&mode=form&tab=core&id=d6efc2cdb7e 6d9e674f4ed558129c071 The Center for Disease Control and Preve ntion [CDC]. (2012). National Public Healt h Performance Standards Program (NPHPSP). Retrieved from http://www.cdc.gov/nphpsp/essentialservices.html

PAGE 203

185 Coleman, J. (1988 ) Social Capital in the Creation of Human Capital. The American Journal of Sociology 94 S95 S120. Colorado C epartment of Education [CDE] (2012). Early Childhood Councils. Retrieved from http://www.cde.state.co.us/early/ECC.htm Colorado Department of Human Services [CDHS]. (2012). Early intervention, the Office of Children, Youth and Families (OCYF). Retrieved from http://www.colorado.gov/ cs/Satellite/CDHS childYouthFam/CBON/1251579373436 Colorado Department of Public Health and Environment [CDPHE]. (2012a). Maternal and child he alth. Retrieved from http://www.colorado.gov/cs/Satellite/CDPHE PSD/CBON/1251617572212 Colorado Department of Public Health and Environment [CDPHE]. (2012b). Health care program for children with special needs (HCP). Retrieved from http://www.colorado.gov/cs/Satellite?c=Page&childpagename=CDPHE PSD%2FCBONLayout&cid=1251617590646&pagename=CBONWrapper Colorado Multiple Institutional Review Board [COMIRB]. (2012). Instructions for new submissions. Retrieved from http://www.ucdenver.ed u/academics/research/AboutUs/comirb/Submissions and Training Portal/Instructions%20For%20New%20Submissions/Pages/Instructions For New Submissions.aspx Conklin, J., Lusk, E., Harris, M., & Stolee, P. (2013). Knowledge brokers in a knowledge network: the ca se of Seniors Health Research Transfer Network knowledge brokers. Implementation Science 8 (1), 1 10. Consortium from Altarum Institute; Centers for Disease Control and Prevention; Robert Wood Johnson Foundation; National Coordinating Center for Public He alth Services and Systems Research. (2012). A national research agenda for public health services and systems. American Journal of Preventive Medicine, 42 (5), S72 78. Creswell, J.W. ( 2007 ) Qualitative inquiry & research design (2nd ed.). Thousand Oaks, CA: Sage Publications, Inc. Cros by, B. C., & Bryson, J.M. ( 2005 ) Leadership for the common g ood: Tackling public p roblems in a s hared p ower w orld San Francisco: Jossey Bass. Currie, J., Stabile, M., Manivong, P., & Roos, L. L. (2010). Child health and young adult outcomes. Journal of Human Resources, 45 (3), 517 548.

PAGE 204

186 Detels, R., Beaglehole, R., Lansang, M.A. & Gulliford, M. (2011). Oxford textbook of public health New York, NY: Oxford University Press. Dhanaraj, C. & Parkhe, A. (2006). Orchestrating innovation networks. Academy of Management Review, 31( 3), 659 669. The Early Childhoo d Technical Assistance Center [ ECTAC]. (2008). Early intervention program for infants and toddlers with disabilities (Part C of IDEA). Retrieved from http://ectacenter.o rg/partc/partc.asp Early Intervention Colorado. (2011). Monitoring reports, plans of correction, public performance and determinations. Retrieved from http://www.eicolorado.org/index.cfm?fuseaction=Documents.content&linkid=276 Early Intervention Colorad o. (2012). State performance plan: Federal IDEA, Part C. Retrieved from http://www.eicolorado.org/index.cfm?fuseaction=Documents.content&linkid=396 Faure, A. M. (1994). Some methodological problems in comparative politics. Journal of Theoretical Politics 6 (3), 307 322. Freeman L C (1979). Centrality in Social Net works: Conceptual Clarification Social Networks 1 215 239. Funnell, S. (2000). Developing and using a program theory matrix for program evaluation and performance monitoring. In Rogers, P., Hacsi, T., Petrosino, A., & Huebner, T. (Eds.), Program theory in evaluation: Challenges and opportunities, new directions for evaluation (pp.91 101). San Francisco, CA: Jossey Bass. Gage, R. ( 1984 ) Federal regional councils: Networking or ganizations for policy management in the intergovernmental system Public Administration Review 44 (2), 134 45. George, G., Zahra, S. A., Wheatley, K. K., & Khan, R. (2001). The effects of alliance portfolio characteristics and absorptive capacity on performance: A study of biotechnology firms. Th e Journal of High Technology Management Research, 12 (2), 205 226. Gerring, J. ( 2007 ) Case study research: Principals and practice New York, NY: Cambridge University P ress. Gillies, D.A. (1982). Nursing management a systems approach Philadelphia: W. B. Saunders Company. George, A., & Bennett, A. ( 2005 ) Case studies and theory development in the social sciences Cambridge, MA: M IT Press.

PAGE 205

187 Gould, J. & Fernandez, J. ( 1989 ) Structures of mediation: A formal approach to brokerage in transaction networks. Sociological Methodology 19 89 126. Granovetter, M. ( 1973 ) The Strength of Weak Ties. American Journal of Sociology 78 (6), 1360 13 80. Granovetter, M. (1985). Economi c action and social structure: T he problem of embeddedness. American Journal of S ociology 91 (3), 481 510. Granovetter, M. (1992). The sociological and economic approaches to labor market analysis: a social structural view. In Granovetter, M. & Swedberg, R. The sociology of economic life ( pp. 233 63 ) New York, NY: Westview Press. Gray, B. ( 1989 ) Collaborating: Finding common groun d for multiparty p roblems San Francis co: Jossey Bass Gray, B ( 2000 ) Assessing inter organizational collaboration: multiple conceptions and multiple methods. In Faulkner, D & de Rond, M. (Eds.) Cooperative strategy: Economic, business, and organizational issues (pp. 243 2 60 ) New York : Oxford University Press. Gray, B. & Wood, D. ( 1991 ) Collaborative a lliances: Moving from practice to theory Journal of Applied Behavioral Science 27 (2) 3 22. Gulati, R., Nohria, N. and Zaheer, A. (2000). Strategic networks. Strategic Management Journal 21 : 203 215. Hall, T. E. & O'Toole, L.J. ( 2004 ) Shaping formal networks through the r egulatory p rocess. Administration & Society 36 (2) 186 207. H anneman, R., & Riddle, M. ( 2005 ) Introduction to social network methods Retrieved from http://faculty.ucr.edu/~hanneman/ Hansen, M. T. (1999). The search transfer problem: The role of weak ties in sharing knowledge across organization subunits. Administrative Science Quarterly, 44 82 111. Haskins, R ( 1989 ) Beyond Metaphor: The Efficacy of Early Childhood Education. American Psychologist 44 (2) 274 82. Hayajneh, Y. (2007). Systems and systems theory. Retrieved from www.hayajneh.org/a/readings/Systems Theory.pdf Hebbeler, K., Spiker, D., Bailey, D., Scarborough, A., Mallik, S., Simeonsson, R., et al. (2007). Early intervention for infants and toddlers with disabilities and their

PAGE 206

188 families: Participants, services, and outcomes. Final report of the national early intervention longitudinal s tudy Ret rieved from http://www.sri.com/neils/pdfs/NEILS_Report_02_07_Final2.pdf Hess, C. & Van Landeghem, K. (2005). Children's mental health: An overview and key considerations for health care system stakeholders. National Institute for Health Care Management, Washington, DC. Retrieved from http://www.nihcm.org/CMHReport FINAL.pdf. Hill, C. (2002). Network literature r eview: Southern Alberta Child and Youth Health Network. Retrieved from www.calgaryhealth.ca Huxham, C & Vangen, S. ( 2005 ) Managing to collaborate: The t heory and p ractice of collaborative a dvantag e. New York NY : Routledge. Hyde, J. K., & Shortell, S. M. (2012). The structure and organization of local and state public health agencies in the US: A systematic review. American Journal of P reventive Medicine, 42 (5), S 29 S41. Institute of Medicine [IOM]. (1996). Primary Care: America's Health in a New Era. Washington, DC: The National Academies Press Institute of Medicine [IOM]. (1998). The future of public health. Washington, DC: The National Academies Press. action and accountability. Washington, DC: The National Academies Press. Institute of Medicine [IOM]. ( 2012a ) Primary care and public heal th: Exploring integration to improve population health Washington, DC: The National Academies Press. Institute of Medicine [IOM]. ( 2012 b ) future Washington, DC: The National Academies Press. Jarillo J. ( 1993 ). Strategic networks: Creating the borderless organization London: Butterworth Heineman. Jones, S. M., & Zigler, E. (2002). The Mozart effect: Not learning from history. Journal of Applied Developmental Psychology 23 (3), 355 372. Kast, F. & Rosenzweig, J. (1972). General systems theory: Applications for organization and management. The Academy of Management Journal, (4) 15: 459. Kaye, N. (2006). healthy mental development: e arly accomplishments and lessons learned from a

PAGE 207

189 five state consortium National Academy for State Health Policy Retrieved from http://www.nashp.org/publication/improving delivery health care supports young childrens healthy mental development Kaye, N., May, J., & Abrams, M. (2006). State policy options to improve delivery of child development services: Strategies from the eight ABCD States National Academy for State Health Policy Retrieved from http://www.nashp.org/publication/state policy options impr ove delivery child development services strategies eight abcd Kaye, N., & Rosenthal, J. (2008). Improving the delivery of health care that supports learned from a five state co nsortium (updated) National Academy for State Health Policy Retrieved from http://www.nashp.org/publication/improving delivery health care supports young childrens healthy mental development. Kilduff, M. & Tsai, W. (2003). Social networks and organizati ons Sage Publications. King, T.M., & Glascoe, F.P. (2003). Developmental Surveillance of Infants and Young Children in Pediatric Primary Care. Current Opinions in Pediatrics, 15 (6), 624 29. Krackhardt D. (1994). Graph theoretical dimensions of informal organizations. Computational Organization T heory, 89 (112), 123 140. Krackhardt, D. (1992). The strength of strong ties: The importance of philos in organizations. In Nohria, N. & Eccles, R.G., Networks and organizations: Structure, form, and action (pp.216 239). Boston, MA: Harvard Business School Press. Krackhardt, D. (2010). Social networks. Encyclopedia of group processes and intergroup relations 817 821, Kuh, D., & Ben Shlomo, Y. (2004). A life course approach to chronic disease epidemiology (Vol. 2). New York, NY: Oxford University Press. Lawless, M. W., & Moore, R. A. (1989). Interorganizational systems in public service delivery: A new application of the dynamic network framework. Human Relations, 42 (12), 1167 1184. Lemaire, R. H., & Provan, K. G. (2009). Network governance in a publicly funded child and youth health network: Sub network embeddedness, cohesiveness, and the role of brokers. In Public Management Research Conference, Ohio State, October. Lilienfeld, R. (1978). The rise of systems t heory : An ideological a nalysis New York NY : John Wiley and Sons

PAGE 208

190 Long, W. E., Bauchner, H., Sege, R. D., Cabral, H. J., & Garg, A. (2012). The value of the medical home for children witho ut special health care needs. Pediatrics 129(1), 87 98. Lowe, J. W. (1985). Qualitative systems theory: its utility and limitations. Journal of Anthropological Research 42 61. Lundin, M. (2007). Explaining cooperation: How resource interdependence, goal congruence, and trust affect joint actions in policy implementation. J ournal of Public Administration Research and Theory, 17 (4), 651 672. Mabry, P. L. (2011). Making sense of the data explosion: The promise of systems science. American Journal of Preventive Medicine 40 (5 S2): S159 S161 Mabry, P. L., Olster, D. H., Morgan, G. D., & Abrams, D. B. (2008). Interdisciplinarity and systems science to improve population health: a view from the NIH Office of Behavioral and Social Sciences Research. American Journal of Preventive Medicine 35(S2), S211. Mackrides, P. S., & Ryherd, S. J. (2011). Screening for developmental delay. American Family Physician 84 (5), 544. Masseurs, J. (2012). Understanding network effectiveness: An in depth analysis of structural and contextual influences on the performance of inter organizational networks Retrieved from http://arno.uvt.nl/show.cgi?f id=127316 Maternal and Child Health Bureau [MCHB]. (2012a). Title V Mate rnal and Child Health services block grant p rogram. Retrieved from http://mchb.hrsa.gov/programs/titlevgrants/ Maternal and Child Health Bureau [MCHB]. (2012b). About Us. Retrieved from http://mchb.hrsa.gov/about/index.html Mays GP, McHugh MC, Shim K, et al. (2006). Institutional and economic determinants of public health system performance. American Journal of Public Health 96 (3), 523 5 31. McGuire, M. ( 2006 ) Collaborative public management: Assessing what we know and how we know it. Public Administration Review 66 (s1), 33 47. ( 2001 ) Managerial strategies and behavior in networks: A model with evidence from U.S. public education. Journal of Policy Analysis and Management 11 (3) 271 93. Mental Health Center of Denver [MHCD] (2012). What We Do. Retrieved from

PAGE 209

191 http://mhcd.org/what we do Merriam, S. B. ( 2009 ) Qualitative research: A guide to design and implementation San Francisco, CA: Jossey Bass. Merrill, J., Caldwell, M., Rockoff, M. L., Gebbie, K., Carley, K. M., & Bakken, S. (2008). Findings from an organizational network analysis to support local public health management. Journal of Urban Health, 85 (4), 572 58 4. Miles, R. & Snow, C. (1986). Organizations: New concepts for new forms. California Management Review, 28 (3), 62 73. Milwar d, H. B. & Provan, K. G. ( 1998 ) Measuring n etwork s tructure. Public Administration 76 (2) 387 407. Mitchell, S. M., & Shortell, S. M. (2000). The governance and management of effective community health partnerships: A typology for research, policy, and practice. Milbank Quarterly, 78 (2), 241 289. Monge, P. R., & Contractor, N. S. (2003). Theories of communication networks (pp. 141 223). New York: Oxford University Press. Moreno, J. L. (1934). Who shall survive?: A new approach to the problem of human interrelations Nervous and mental disease monograph series, no 58 Washington, DC: Nervous and Mental Disease Publ ishing Co. Moynihan, D. P., Provan, K. G., & Lemaire, R. H. 2012. Core c oncepts and k ey ideas for u nderstanding p ublic s ector o rganizational n etworks: Using r esearch to i nform s cholarship and p ractice. Public Administration Review 72 (5), 638 648. Nadel S.F. ( 1957 ) The theory of social structure. London: Cohen and West. National Association of County and City Health Officials [NACHO]. (2010). Public health infrastructure and systems. Retrieved from http://www.naccho.org/topics/infrastructure/ National A cademy for State Health Policy [ NASHP ] (2012). About ABCD. Retrieved from http://www.nashp.org/abcd history National Cancer Institute. (2007, April) Greater than the sum: Systems thinking in tobacco control. Tobacco control monograph no. 18 Bethesda, MD: U.S. Department of Health and Human Services, National Institutes of Health, National Cancer Institute. NIH Pub. No. 06 6085.

PAGE 210

192 National Early Childhood Technical Assistance Center [NICHY]. (2012). Effective Practices in Early Intervention. Retrieved from http://nichcy.org/babies/effectivepractices National Public Health Performance Standards Program [NPHPSP]. (2007). Local public health system assessment instrument Retrieved from http://www.cdc.gov/nphpsp/documents/local/local.bookleta. pdf Neely, A., Gregory, M., & Platts, K. (1995). Performance measurement system design: a literature review and research agenda. International Journal of Operations & Production Management 15 (4), 80 116. Nelson, C. S., Wissow, L. S., & Cheng, T. L. (2003). Effectiveness of anticipatory guidance: recent developments. Current Opinion in Pediatrics, 15 (6), 630 635. Organization for Economic Cooperation and Development [OECD]. (2009). OECD factbook 2009: Economic, environmental and social statistics OECD Publishing. doi: 10.1787/factbook 2009 en Based Agendas in Public Administration. Public Administration Review, 57 (1), 45 52. O'Too le, L. J., & Meier, K. J. ( 2004 ) Public m anagement in i ntergovernmental n etworks: Matching s tructural n etworks and m anagerial n etworking. Journal of Public Administration Research and Theory 14 (4) 469 494. Provan, K. G., Fish, A., & Sydow, J. (2007). Interorganizational networks at the network level: A review of the empirical literature on whole networks. Journal of management 33 (3), 479 516. Provan, K. G., & Lemaire, R. H. (2012). Core concepts and key ideas for Uunderstanding public sector organizational networks: Using research to inform scholarship and practice. Public Administration Review 72 (5), 638 648. Provan, K. G. & Milward, B. ( 1991 ) Institutional level norms and organizational involvement in a service implementati on network. Journal of Public Administration Research and Theory, 1 (4) 391 417. Provan, K.G., & Milward, B (1995). A preliminary theory of interorganizational network effectiveness: A comparative study of four community mental health systems Administra tive Science Quarterly, 40 (1) 1 33. Provan, K.G., & Milward, B ( 2001 ) Do networks really work? A framework for evaluating public sector organizational networks. Public Administration Review, 61 (4) 414 23.

PAGE 211

193 Provan, K.G; Veazie, M.A.; Staten, L.K. & Teufel Shone, N.I. ( 2005 ) The u se of n etwork a nalysis to s trengthen c ommunity p artnerships. Public Administration Review, 65 (5) 603 613 Przeworski, A. & Teune, H. ( 1970 ) The logic of comparative social inquiry New York: Wiley. Rethemeyer, R. K. ( 200 5 ) Conceptualizing and m easuring c ollaborative n etworks. Public Administration Review 65 (1) 117 21. Ring, P. S., & Van de Ven, A. H. (1994). Developmental processes of cooperative interorganizational relationships. Academy of Management Review, 90 118. Rosenheck, R., Morrissey, J., Lam, J., Calloway, M., Johnsen, M., Goldman, H., et al. (1998). Service system integration, access to services, and housing outcomes in a program for homeless persons with severe mental illness. American Journal of Public He alth 88(11), 1610 1615. Roussos, S. T., & Fawcett, S. B. (2000). A review of collaborative partnerships as a strategy for improving community health. Annual Review of Public H ealth, 21 (1), 369 402. Russ, S., Garro, N., & Halfon, N. (2010). Meeting children's basic health needs: From patchwork to tapestry. Children and Youth Services Review 32 (9), 1149 1164. Rydz, D., Shevell, M.I., Majnemer, A., & Oskoui, M. (2005). Developmental Screening. Journal of Child Neurology 20 (1) 4 21. Sandstrm, A., & Carlsson, L. (2008). The performance of policy networks: the relation between network structure and network performance. Policy Studies Journal 36 (4), 497 524. Scott, J. ( 2000 ) Social network analysis: A handbook (2nd e d .) Newberry Par k, CA: Sage Publications. Scutchfield, F. D., Prez, D. J., Monroe, J. A., & Howard, A. F. (2012). New Public Health Services and Systems Research Agenda. American Journal of Preventive Medicine 42 (1), 1. Shonkoff, J. P., & Phillips, D. A. (Eds.). (2000). From neurons to neighborhoods: The science of early childhood development Washington DC: National Academies Press. Sices, L. (2007). Developmental screening in primary care: the effectiveness of current practice and recommendations for improveme nt New York, NY: The Commonwealth Fund.

PAGE 212

194 Snyder, E. H., Lawrence, C. N., & Dodge, K. A. (2012). The impact of system of care support in adherence to wraparound principles in Child and Family Teams in child welfare in North Carolina. Children and Youth Services Review 34(4), 639 647. Starfield, B., & Horder, J. (2007). Interpersonal continuity: old and new perspectives. The British Journal of General Practice 57 (540), 527. Strickland, B. B., Jones, J. R., Ghandour, R. M., Kogan, M. D., & Newacheck, P. W. (2011). The medical home: health care access and impact for children and youth in the United States. Pediatrics 127(4), 604 611. Strickland, B. B., Singh, G. K., Kogan, M. D., Mann, M. Y., van Dyck, P. C., & Newacheck, P. W. (2009). Access to the m edical home: new findings from the 2005 2006 National Survey of Children with Special Health Care Needs. Pediatrics 123(6), e996 e1004. Stroul, B. A., Blau, G. M., & Friedman, R. M. (2010). Updating the system of care concept and philosophy Washington, DC: Georgetown University Center for Child and Human Development, National Technical Assistance Center for Stroul, B. A., & Friedman, R. M. (1986). A system of care for severely emotionally disturbed children and youth Washingto n, DC: Georgetown University Child Development Center. Stroul, B. A., & Friedman, R. M. (1994). A system of care for children and youth with severe emotional disturbances Washington, DC: Washington, DC: Georgetown University Child Development Center. Su en J Christenson G M Cooper A Taylor M. (1995). Analysis of the current status of public health practice in local health departments. American Journal of Preventative Medicine, 11 (6S), S51 S54. Sydow, J. ( 1998 ) Organizing and evaluating interfirm networks: A structurationist perspective on netw ork processes and effectiveness Organization Science, 9 (3) 265. Thomson, A. M. & Perry, J. L. ( 2 006 ) Collaboration p rocesses: Inside the b lack b ox. Public Administration Review 66 ( s 6) 20 31. Thomson, A.M., Perry, J.L. & Miller, T.K. (2008 a ). Conceptualizing and measuring collaboration. Journal of Public Administration Research and Theory, 19 (1), 23 56.

PAGE 213

195 Thomson, A. M., Perry, J. L, & Miller, T.K ( 2008 b ) Linking collaboration processes and outcomes. In Bingham, L. B & O'Leary, R.N. Big Ideas in Collaborative Public Management (pp. 97 120) Armonk, New York : M.E. Sharpe. United States Census Bureau. (2010). 2010 Census Data accessed at http://www.census.gov/2010census/ United States Department of Education. (201 2 ). Office of Special Education Programs, Data Accountability Center (DAC) : Infants and Toddlers Receiving Early Intervention Service s in Accordance with Part C 201 2 Retrieved from https://www.ideadata.org/arc_toc13.asp#partcCC United States Department of Health an d Human Services [HHS] (2010). Open government p lan. Retrieved from http://www.hhs.gov/open/plan/opengovernmentplan/openplanversion1_1.pdf United States Department of Health and Human Services [HHS]. (2012).Message from the secretary. Retrieved from http://www.hhs.gov/secretary/about/message.html United States Department of Health and Human Services; United States Department of Education; United States Department of Justice. (2000). Report of the Surgeon General's conference on children's mental health: A national action agenda Washington (DC). Retrieved from http://www.ncbi.nlm.nih.gov/books/NBK44233/ United States Go vernment Accountability Office [GAO] (2012). 2012 Performance and Accountability Report. Retrieved from http://www.gao.gov/about/perfmeasures.html Uzzi, B. 1996. The sources and consequences of embeddedness for the economic performance of organizations: The network effect. American Sociological Review, 61 : 674 698. Uzzi, B. 1997. Social str ucture and competition in inter firm networks: The paradox of embeddedness. Administrative Science Quarterly, 42 : 35 67. Uzzi, B. 1999. Embeddedness in the making of financial capital: How social relations and networks benefit firms seeking financing. American Sociological Review, 64 : 481 505. Varda, D. ( 2012 a ) PARTNER survey questions. Retrieved from http://www.partnertool.net/resources. Varda, D. (2012b). PARTNERTool. [ Computer program]. Available from http://www.partnertool.net/.

PAGE 214

196 Varda, D. M., Chandra, A., Stern, S. A., & Lurie, N. ( 2008 ) Core dimensions of connectivity in public health collaboratives. Journal of Public Health Management Practice, 14 (5) E1 E7 Veazie, M. A., T eufel Shone, N.I., et al. ( 2001 ) Building community capacity in Public Health: The role of action oriented partnerships. Journal of Public Health Management and Practice 7 (2) 21 32. Von Bertalanffy, L. (1972). The History and Status of General Systems Theory. Academy of Management Journal 15(4), 407 426 Wasserman, S., & Galaskiewicz, J. (Eds.). (1994). Advances in social network analysis: Research in the social and behavioral sciences Sage Publications, Inc Wasserman, S. & Faust K ( 1994 ) Social network analysis: Methods and applications Cambridge, United Kingdom: Cambridge University Press. Weiner B., Alexander & Zuckerman. ( 2000 ) Strategies for effective management participation in community health partnerships. Health Care Management Review 25 (3): 48 66. Wellman, B, 1988. Structural a nalysis: From method and metaphor to theory and substance. I n Wellman, B., & Berkowitz, S. D. ( Eds. ) Social structures: A network approach (p p 19 61 ) Cambridge: Cambridge University Press. Wise P. & Blair M. (2007). The UNICEF Report on child well being: a commentary. Academic Pediatrics 7 265 267 Whitchurch, G. G., & Constantine, L. L. (1993). Systems theory. In P. G. Boss, W. J. Doherty, R. LaRossa, W. R. Schumm, & S. K. Steinmetz (Eds.), Sourcebook of family theories and methods: A contextual approach (pp. 325 352). New York: Plenum Press. Wholey, D. R., Gregg, W., & Moscovice, I. (2009). Public health systems: a social networks perspective. Health Services R esearch, 44 (5p2), 184 186 Wood, D. J. & Gray, B ( 1991 ) Toward a comprehensive theory of collaboration. Journal of Applied Behavioral Science 27 (2) 139 162. Yin, R. K. ( 2003 ) Case study research: Design and methods Thousand Oaks, CA: Sage Publications, Inc. Zaheer, A., McEvily, B., & Perrone, V. (1 998). Does trust matter? Exploring the effects of interorganizational and interpersonal trust on performance. Organization S cience 9 (2), 141 159.

PAGE 215

197 Zahner, S.J. ( 2005 ) Local public health system partnerships. Public Health Reports 120 (1) 76 83. Zahner, S.J. Kai ser, B., & Kapelke Dale, J. ( 2005 ) Local partnerships for community assessment and planning. Jour nal of Public Health Management and Practice, 11 (5) 460 464. Ziviani, J., Darlington, Y., Feeney, R., & Head, B. (2011). From policy to practice: A program logic approach to describing the implementation of early intervention services for children with physical disability. Evaluation and Program Planning 34 (1), 60 68. Zuckerman, B. Aug ustyn, M. & P arker S ( 2001 ) Child development in pediatrics: beyond rhetoric. Archives of Pe diatric and Adolescent Medicine, 155 1294 12 95.

PAGE 216

198 APPENDIX 1. Early Intervention Process in Colorado SOURCE: Early Intervention Colorado, 2012

PAGE 217

199 APPENDIX 2. Survey Invitation Your organization has been identified by the Assuring Better Child health Development (ABCD) State team as a community stakeholder working to ensure standardized developmental screening, referral and follow up for children your community. Thank you for al l your work towards this important goal. Given your involvement in this work, your organization is being asked to be in a research study titled: Partners for Healthy Development Study Purpose This study aims to learn more about the social structure of interorganizational partnerships in public health, specifically asking how different characteristics and attributes of partnerships affect performance. This study focuses on the relationships betw een organizations working together towards the same goal: working to ensure standardized developmental screening, referral and follow up for children in their community. A total of six stakeholder groups will be in this study, including all of the organiz ations that are a part of each stakeholder group. Partnerships and collaboration are inherent in the public health system where public, private and nonprofit organizations work together to assure the conditions for population health. This study aims to arm you with the knowledge to prompt strategic and thoughtful discussion to foster stronger inclusion, involvement and commitment where it most makes sense and will have the greatest impact to ensure standardized developmental screening, referral and follo w up for children your community. More broadly, this study aims to provide insight into the dimensions of collaboration, adding to a much needed knowledge base to guide the use, formation and management of partnerships and collaboration for a strong publi c health system. Your Participation You are being asked to complete an online survey. This survey will include 20 questions, and will take about 20 minutes. It will list all the organizations identified by ABCD as working to ensure standardized developmental screening, referral and follow up for children in your community, and request that you indicate your involvement and relationship with the other organizations on the list. Taking part in this study is voluntary. If you choose to take part, y ou have the right to skip questions or stop at any time. Confidentiality and Anonymity No personal information will be collected. All questions are strictly about organizations (not individuals). This means that as a respondent to the survey, you are asked to represent your organization. The data will only ever be presented as aggregate scores (combined average of all survey respondents within your community) and at the network level (mathematical calculations about the stakeholder group as a whole). Analysis of the data will be presented as part of a research dissertation in written and oral form, but will

PAGE 218

200 (This will always remain confidential). Aggregate findi ngs will also be presented to you and other organizations in your community that have taken the survey, and to the Colorado ABCD State Team The research data and notes will remain confidential. (All survey results, analysis documents and research notes will be password protected and kept on an encrypted computer or locked filing cabinet). Funding for the Study This study is part of a dissertation for a PhD in Public Affairs at the University of Colorado Denver School of Public Affairs by Robyn Mobbs. The National Coordinating Center for Public Health Systems and Services Research, with funding from the Robert Wood Johnson Foundation, has awarded Robyn the Assuring the Future of Public Health Systems Research: Dissertation Award to support this resear ch. Contact Details The researcher leading this study is Robyn Mobbs. If you have questions or concerns, you may call Robyn at 303 204 0640. You can also call the Colorado Multiple Institutional Review Board (COMIRB) at (303) 724 1055.

PAGE 219

201 APPENDIX 3. Su rvey Questions Partners for Healthy Development: A network analysis of community stakeholders working to ensure standardized developmental screening, referral and follow up for children in six Colorado communities. 1: Please select your organization/program from the list < local organizations will be listed by name for each of the following categories (more than one organization per category is OK > : Behavioral health Child care provider Child Find/ BOCES Health Department Head Start Health care provider Home visitation Early Childhood Council Early Intervention Family Resource Center Parent/family member Religious Organization Service Provider 2: How is your organization/program participating to ensu re standardized developmental screening, referral and follow up for children in your community? 3: How long have you been participating to ensure standardized developmental screening, referral and follow up for children in your community (in months)? 4: Please indicate what your organization/program, or can potentially contribute, to standardized developmental screening, referral and follow up for children in your community. (Please choose all that apply). Funding In kind resources. If you are providing in kind resources. Please define. Paid staff Volunteers / volunteer staff Data resources, including data sets, collection and analysis Information/feedback Specific health expertise Expertise other than health Community connections

PAGE 220

202 Fiscal management Facilitation/leadership developmental screening, referral and follow up for children in your community? Funding In kind resources. If you are providing in kind resources. Please de fine. Paid staff Volunteers / volunteer staff Data resources, including data sets, collection and analysis Information/feedback Specific health expertise Expertise other than health Community connections Fiscal management Facilitation/leadership 6: Please indicate which of the following outcomes have been achieved as a result of the focus on standardized developmental screening, referral and follow up for children in your community. (Please choose all that apply). Clearer understanding of the developmenta l and behavioral needs of children Clearer understanding of what is needed to better meet those needs Promotion and formation of working relationships between health care practices and state and local programs, services, and resources, including Community Centered Board (by name), Child Find/BOCES Increase in the number of health care practices that have implemented standardized developmental screening and referral Promotion of early identification and referral Improved referral follow through process Incre ase in the number of referrals Increased resource sharing Increased knowledge sharing Stronger link with local Community Center Board point of view which of the above describes the most successful outcome of the focus on standardized developmental screening, referral and follow up for children in your community? (Please choose one). Clearer understanding of the developmental and behavioral needs of children Clearer understanding of what is needed to better me et those needs Promotion and formation of working relationships between health care practices and state and local programs, services, and resources, including local Community Centered Board (by name), Child Find/BOCES Increase in the number of health care practices that have implemented standardized developmental screening and referral

PAGE 221

203 Promotion of early identification and referral Improved referral follow through process Increase in the number of referrals Increased resource sharing Increased knowledge sha ring Stronger link with local Community Center Board 8: From your point of view, how successful has this focus been for children in your community? Not Successful Somewhat Successful Successful Very Successful Completely S uccessful 9: What aspects of collaboration contribute to this success? (Please choose all that apply). Bringing together diverse stakeholders Meeting regularly Exchanging information/knowledge Sharing resources Formation of collaborative relationships Collective decision making Emphasis on personal relationships Shared Funding Shared Expertise Other Network Questions: 10: From the list, please select the organizations/programs with which you have an established relationship (either formal or informal). In subsequent questions you will be asked about your relationships with these organizations/programs in the context of standardized developmental screening, referral and follow up for children in your community < local organizations will be listed by name from the following sectors > : Behavioral health Child care provider Child Find/ BOCES Health Department Head Start Health care provider Home visitation Early Childhood Council Early Intervention Family Resource Center Parent/family member

PAGE 222

204 Religious Organization Service Provider 11: At this point in time, how frequently does your organization/program work with this organization/program on issues related to standardized developmental screening, referral and follow up for children in your community? Never/We only interact on issues unrelated to standardized developmental screening, referral and follow up Once a year or less Every few months Every month Every few weeks Every week hip with this organization/program entail (in the context of standardized developmental screening, referral and follow up for children in your community)? None Cooperative Activities: such as exchanging information, closing the loop on referrals, attendin g meetings together, and offering resources to partners. Coordinated Activities: Include cooperative activities in addition to intentional efforts to enhance each other's capacity for the mutual benefit of programs. (Example: Separate granting programs utilizing shared administrative processes and forms for application review and selection.) Integrated Activities: In addition to cooperative and coordinated activities, this is the act of using commonalities to create a unified center of knowledge and pro gramming that supports work in related content areas. (Example: Developing and utilizing shared priorities for funding effective prevention strategies. Funding pools may be combined.) power and influence (i n the context of standardized developmental screening, referral and follow up for children in your community)? *Power/Influence: The organization/program holds a prominent position in the community be being powerful, having influence, success as a change agent, and showing leadership Not at all A small amount A fair amount A great deal level of involvement (in the context of standardized developmental screening, referral and follow up for children in your community)? Level of Involvement: The organization/program is strongly committed and active in the partnership and gets things done. Not at all

PAGE 223

205 A small amount A fair amount A great deal resource contribution (in the context of standardized developmental screening, referral and follow up for children in your community)? *Contributing Resources: The organization/program brings resources to the partnership like funding, information, or other resources. Not at all A small amount A fair amount A great deal 16: How reliable is the organization/program (in the context of standardized developmental screening, referral and follow up for children in your community)? *Reliable: this organization/pr ogram is reliable in terms of following through on commitments. Not at all A small amount A fair amount A great deal 17: To what extent does the organization/program share a common vision of standardized developmental screening, referral and follow up fo r children in your community ? *Mission Congruence: this organization/program shares a common vision of the end goal of what working together should accomplish. Not at all A small amount A fair amount A great deal 18: How open to discussion is the organ ization/program (in the context of standardized developmental screening, referral and follow up for children in your community)? Open to Discussion: this organization/program is willing to engage in frank, open and civil discussion (especially when dis agreement exists). The organization/program is willing to consider a variety of viewpoints and talk together (rather than at each other). You are able to communicate with this organization/program in an open, trusting manner. Not at all A small amount A fair amount A great deal

PAGE 224

206 19: To the best of your knowledge, how would your organization define success of this focus on standardized developmental screening, referral and follow up for children in your community? 20: To the best of your knowledge, how would your organization describe the benefits of participating in this effort?

PAGE 225

207 APPENDIX 4. Stakeholder Network Self Assessment Responses Figure A 1 Self Assessment for Stakeholder Network Alpha. Figure A 2 Self Assessment for Stakeholder Network Bravo. Figure A 3 Self Assessment for Stakeholder Network Charlie. 0 2 4 6 8 Not Successful Somewhat Successful Successful Very Successful Completely Successful Alpha Self Assessment Responses 0 2 4 6 8 Not Successful Somewhat Successful Successful Very Successful Completely Successful Bravo Self Assessment Responses 0 1 2 3 4 5 Not Successful Somewhat Successful Successful Very Successful Completely Successful Charlie Self Assessment Responses

PAGE 226

208 Figure A 4 Self Assessment for Stakeholder Netwo rk Delta. Figure A 5 Self Assessment for Stakeholder Network Echo. 0 1 2 3 4 5 Not Successful Somewhat Successful Successful Very Successful Completely Successful Delta Self Assessment Responses 0 1 2 3 4 5 6 Not Successful Successful Completely Successful Echo Self Assessment Responses