Citation
Assessing the risk of homeless adolescents becoming homeless adults

Material Information

Title:
Assessing the risk of homeless adolescents becoming homeless adults
Creator:
Discenza, Suzanne Shepherd
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
xvi, 258 leaves : ; 28 cm

Thesis/Dissertation Information

Degree:
Doctorate ( Doctor of Philosophy)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
School of Public Affairs, CU Denver
Degree Disciplines:
Public Affairs
Committee Chair:
Beatty, Kathleen
Committee Members:
Burton, Lloyd
Dodge, Mary
Scandlyn, Jean
Hagedorn, Susan

Subjects

Subjects / Keywords:
Homeless youth ( lcsh )
Homelessness ( lcsh )
Risk assessment ( lcsh )
Homeless youth ( fast )
Homelessness ( fast )
Risk assessment ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 252-258).
General Note:
School of Public Affairs
Statement of Responsibility:
by Suzanne Shepherd Discenza.

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:
63803632 ( OCLC )
ocm63803632
Classification:
LD1193.P86 2004d D57 ( lcc )

Full Text
ASSESSING THE RISK OF HOMELESS ADOLESCENTS
BECOMING HOMELESS ADULTS
Suzanne Shepherd Discenza
B.S., Oklahoma State University, 1971
M.S., University of Oklahoma, 1974
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Public Affairs
2004
by
\
r


2004 by Suzanne Shepherd Discenza
All rights reserved.


This thesis for the Doctor of Philosophy
degree by
Suzanne Shepherd Discenza
has been approved
by


Discenza, Suzanne (Ph.D., Public Affairs)
Assessing the Risk of Homeless Adolescents Becoming Homeless Adults
Thesis directed by Professor Kathleen Beatty
ABSTRACT
Homelessness among young people in the United States has been identified as
a serious and growing problem, and yet rigorous research on this population is
considerably sparser than research on homeless adults. Given evidence indicating
that a substantial percentage of homeless adults were first homeless as children, this
exploratory and comparative study addresses why some youths remain homeless as
adults while others are able to permanently exit the streets.
Utilizing qualitative research methodology, primary data collection consisted
of 48 in-depth retrospective interviews of young adults in Denver, Colorado, who had
been homeless as adolescents. While some participants had transitioned into
productive adulthood, others remained on the streets or were in transitional stages. A
modified grounded theory approach served as the principal mode of analysis, drawing
upon theory from public policy and management and from developmental assets
research.
The findings of this study suggest not only that several significant factors
(related to abuse and institutional failure) serve as the most frequent precursors to
initial youth homelessness, but also that additional factors (including substance abuse
and police records) serve to keep youths chronically on the streets. Conversely, a
combination of internal assets (such as resiliency and self-motivation) and external
assets (including satiation of basic needs and appropriate interventions) are key in
enabling youths to permanently exit homelessness. Findings further suggest
differences in the experiences of young homeless females and males, as well as
between homeless homosexual and heterosexual youths.
Policy implications include the identification of how principal findings may
be used to improve advocacy efforts on behalf of this population, to shape best
practice strategies for youth services agencies, and to capitalize on the positive
IV


assets of youths in helping themselves out of homelessness. Reframing this
population as deserving of greater attention by public policy makers is a major
theme of this research.
A principal contribution of this project is posited by listening to the
recommendations of the young adult informants regarding opportunities to aid service
providers in transitioning homeless youths into mainstream adulthood. Involving
them in collaborative research efforts will only serve to strengthen programs on their
behalf.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Signed
Kathleen Beatty
v


DEDICATION
These pages are dedicated to the memory of Dr. Franklin James, superb
mentor and friend, who conceived and supported this qualitative research project to
explore why some homeless youths remain homeless as adults while others are able
to permanently exit the streets. His vision and compassion toward those less
fortunate continued to guide me throughout this study.
Not to forget those who continue to enrich my life with their love and
unequivocal support, I also dedicate this dissertation, with deepest gratitude, to my
husband and soul mate, Richard, our wonderful children and their equally wonderful
spouses, our precious baby grandchildren, and my exceptional parents, C.B. and
Martha B. Shepherd. May this work, in some way, provide a better world for each of
you.


ACKNOWLEDGMENTS
Many exceptional people, almost too numerous to mention, contributed to
making this research effort possible. Special thanks must go to Dr. Kathleen Beatty,
Dean of the Graduate School of Public Affairs and my dissertation committee chair,
who graciously stepped in to direct this project upon the untimely passing of Dr.
Franklin James. Dr. Lloyd Burton not only offered to fill the gap left by Dr. James
absence, but also provided excellent guidance in application of important principles
of public policy and management to ensure focus for this research. Dr. Mary Dodge,
with her special knowledge of juveniles in the criminal justice system, provided fresh
insights and helpful editorial comments as I neared completion of this thesis. I will
be forever grateful to my outside committee members, Dr. Jean Scandlyn and Dr.
Susan Hagedom, for their unwavering encouragement and for the excellent
perspectives they brought to this study from their respective fields of applied
anthropology and nursing. Their expert understanding of homeless adolescents, and
of developmental assets and resiliency theory, particularly informed this research.
This study would not have been possible without the participation and
enthusiastic support of the staff members of Urban Peak, a not-for-profit agency
providing shelter and support services to homeless and runaway youths. I am in
particular dept to the leadership of Roxane White, Jerene Petersen, and James Van
Leeuwen, and to members of the Urban Peak Research Committee, who provided
outstanding critiques of the research design and who put me in contact with many of
the studys participants. Certainly, kudos must go to the young informants who
unconditionally agreed to share their life histories or to participate on the Youth
Advisory Panel. We laughed, cried, and strategized together.
A generous grant from the Fannie Mae Foundation was especially helpful in
supporting this research. My hope is that the fruits of this labor will encourage future
funding of research on behalf of homeless young people.
Finally, I am grateful to many others who provided significant assistance and
encouragement toward completion of this project. I especially want to thank several
GSPA staff members who were always there to provide clerical, scheduling,
technical, and budgetary support, most importantly Annie Nelson, Dawn Savage, and
Kathy Kilpatrick. My cohorts at GSPA and colleagues at the Metropolitan State
College of Denver (Dr. Nancy Shanks deserves special mention) offered skilled
commentary and unparalleled friendship in seeing me through this effort. My
talented student assistant at MSCD, Rachel Miller, provided invaluable assistance in
the development of models and tables for this thesis. Last but not least, I want to
recognize the incredible patience and good will of my personal friends and extended
family. Thank you for always being there.


CONTENTS
Figures.........................................................xv
Tables.........................................................xvi
CHAPTER
1. INTRODUCTION .................................................1
Application of Theory from Public Policy and Public Management.4
Examining and Extending Developmental Assets Theory.......8
Contributions to Theory and Practice.....................10
Overview of the Thesis................................. 12
Risk Factors for Homelessness............................12
Policy Orientation..................................... 13
2. REVIEW OF THE LITERATURE....................................21
Expanding Attention to Adolescent Homelessness...........22
Common Problems Characteristic of Homeless Youth.........24
Risk Factors that Predispose Youths to Becoming Homeless.26
Resiliency in Homeless Adolescents.......................28
On Homeless Youths Becoming Homeless Adults..............29
The Importance of Developmental Assets in Preventing
Future Homelessness................................31
viii
L


Other Factors and Predictors
33
Researcher Collaboration with Homeless Youths..................35
Effects of Gender and Sexual Orientation on Youth
Homelessness...................................................36
Stages of Transition of Homeless Youth.........................39
Best Practices in Interventions for Homeless Adolescents.....41
Summary of Literature Review...................................44
3. DESIGN OF THE RESEARCH.............................................46
Risk Factors for Homelessness..................................47
Policy Orientation.............................................47
On the Importance of Qualitative Research and Life Histories...48
Site of the Study..............................................51
Sample.........................................................53
Access and Sample Selection.................................... 60
Pilot Study.................................................... 64
In-depth Interviews..:......................................... 64
Interviewing Techniques and Parameters...................65
Topics Covered...........................................67
Taping, Transcription, and Preparation of Transcripts
for Analysis.............................................69
Youth Advisory Panel...........................................70
ix


Focus Group of Providers...................................74
Data Analysis..............................................75
Trustworthiness and Credibility............................81
Ethical Concerns...........................................85
Limitations................................................87
Summary of Research Design.................................90
4. CHALLENGING MAINSTREAM DEVELOPMENTAL ASSETS...................91
Theory for Normal Development............................93
Absence of Basic Needs in Homeless Youths..................94
Societal Misperceptions in Holding Homeless Youths Down...102
A Different Set of Assets.................................107
Conclusions Concerning Assets.............................111
5. THE DOWNWARD SPIRAL OF DISABLING CONDITIONS..................113
Family Dysfunction and Abuse..............................113
Physical, Sexual, and Psychological Abuse...........114
Other Family Dysfunction............................118
Failure of Social Institutions............................121
Educational Systems.................................122
Social Services Agencies............................125
x


Policing Agencies and the Criminal Justice System.......129
Effects of Gender and Sexual Orientation.......................136
Specific Gender Issues....................................137
Sexual Orientation Issues...............................140
Physical and Mental Health Issues.............................. 142
Drug and Alcohol Abuse...................................143
Mental Health Issues.....................................147
Physical Health Issues.................................. 151
Models of Homelessness.........................................152
Conclusions Concerning Disabling Conditions......................156
6. THE UPWARD IMPETUS OF ENABLING ASSETS..............................159
Internal Assets.................................................159
Resiliency and Survival Skills...........................161
Marketable Skills........................................165
Core of Human Capital and Values.........................168
Sense of Self and Sense of Agency........................171
Sense of Family..........................................174
Sense of Purpose and Goal Orientation....................177
External Assets.................................................180
Urban Peak...............................................181
xi


Employment.........................................185
Education and Training.............................187
Affordable Housing.................................189
Access to Health Care..............................190
Healthy Relationships and the Special Case of
Street Families....................................192
Concluding Remarks Regarding Enabling Assets..............195
7. RECOMMENDATIONS FROM RESEARCH PARTICIPANTS...................197
Recommendations for Agencies Serving Homeless Youth.......197
Behavioral Issues..................................198
Expansion of Existing Resources....................200
Personal Support............................... .202
Needs and Treatment Oriented Recommendations.......203
Life Skills and Activities.........................204
Aftercare..........................................205
Gender and Sexual Orientation Issues...............206
Recommendations for Public Institutions...................207
Schools, Teachers, and School Counselors...........208
Social Services Agencies...........................210
Juvenile Justice and Criminal Justice Systems......211
Health Care System.................................212
Xll


Recommendations for Families...............................213
Parents, Guardians, and Other Adult Advocates.......213
Recommendations to Siblings.........................215
Recommendations for Homeless Youths........................216
Summary and Implications...................................217
8. SUMMARY AND FUTURE IMPLICATIONS...............................218
Overview of Significant Findings...........................219
Factors Shaping Initial Adolescent Homelessness.....219
Factors Affecting Adult Pathways Into/Out of
Homelessness........................................220
Comparisons of Findings by Gender and Sexual
Orientation.........................................222
Policy and Program Implications............................224
Reframing Social Constructions of Homeless
Adolescents.........................................224
Adopting Best Practice Models for Serving Homeless
Youths..............................................228
Extending Developmental Asset Models and
Capitalizing on Positive Youth Assets...............231
Suggestions for Future Research............................234
Conclusion.................................................236
xm


APPENDIX
A. SOCIAL CONSTRUCTIONS OF TARGET POPULATIONS...237
B. MASLOWS HIERARCHY OF NEEDS.................238
C. PROCHASKAS STAGES OF CHANGE................239
D. CONSENT FORMS FOR INTERVIEWS................240
E. TENTATIVE INTERVIEW GUIDE...................247
F. SAMPLE PAGE OF INTERVIEW TRANSCRIPT.........249
G. GLOSSARY OF STREET TERMS....................250
BIBLIOGRAPHY........................................252
xiv


FIGURES
Figure
4.1 40 Developmental Assets........................
4.2 Extended Model of Assets for Homeless Adolescents
5.1 Pathways to Homelessness.......................
5.2 Vicious Cycles of Homelessness.................


TABLES
Table
3.1 Demographics of Primary Sample................................. 57
3.2 Demographics of Youth Advisory Panels............................ 72
5.1 Patterns of Abuse......................................... 115
5.2 Youth Offenses............................................. 134
5.3 Substance Abuse............................................. 144
6.1 Urban Peak Services Valued by Youths............................ 182
xvi


CHAPTER 1
INTRODUCTION
Homelessness among young people in the United States has been identified
as a serious, complex, and growing problem, and yet rigorous research on this
population is considerably sparser than contemporary research on homeless adults or
families with children. More recent estimates of the prevalence of homelessness
among adolescents include reports that this population may exceed two million
(Rew, 1996) and that from 5 percent (Robertson & Toro, 1999) to 7.6 percent
(Ringwalt et al., 1998) of all youths aged 12 to 17 have experienced at least one
homeless episode. According to Robertson and Toro, these figures appear to place
adolescents at greater risk for literal homelessness than adults (1999, p. 1). In the
city of Denver alone the number of known homeless adolescents on any given day
increased from 200 to 400 youths in the two years between 2001 and 2003, as
reported by a 2003 point-in-time survey by the Colorado Coalition for the Homeless
(Colorado Dept, of Human Services, 2003).
One concern which has been raised is whether adolescent homelessness is a
potential precursor to adult homelessness. The primary purpose and focus of the
current research is to not only address this concern but also to investigate the risk
factors most likely to effect this outcome. As noted by Robertson and Toro,
1


Homelessness itself potentially poses health risks to youths and can interrupt
normal socialization and education, which likely affects a young persons future
ability to live independently (1999, p. 3). Strongly emphasizing the need for more
research on the relationship between homelessness in youths and adult homelessness,
Franklin James refers to evidence that suggests that many of todays homeless
young people will be tomorrows homeless adults unless they get the help they need
(1999, p. 24). At the time of his untimely death in 2001, Dr. James was interested in
supporting a qualitative research project exploring the nuances of why some
homeless youths remain homeless as adults while others are able to permanently exit
the streets.
Given evidence indicating that a substantial percentage of homeless adults
were first homeless as children, ranging from a conservative 9 percent to 26 percent
(Robertson & Toro, 1999, p. 15) to 50 percent or more (James, 1999, abstract), the
importance of this exploratory and comparative research study cannot be overstated.
Obviously, not all homeless adolescents drift into chronic homelessness as adults,
but what factors influence the difference?
Of particular pertinence to the current investigation, in its study of Denver
homeless youths, are unpublished findings from a 1990 Colorado survey showing
that the median age at which homeless adults were first on their own was reported
to be 17 and that by 17 or 18, half of todays homeless adults were already on their
own, frequently with children (James, 1999, pp. 24-25). My own observations as a
2


practitioner and manager in public health and home health care settings provided
innumerable opportunities to hear the stories of system kids (in group homes or
foster care) and homeless adults who had initially run away from abusive home
environments well before the age of 18. These findings illustrate the very direct
linkage between youth homelessness and eventual adult/family homelessness.
Glasser and Bridgman emphasize the need for further research in this area,
citing the particular importance of questions such as, what is the likelihood of a
homeless youth becoming a homeless adult, and, most important, what types of
services and interventions help homeless youth become housed and stay housed?
(1999, p. 23). Providing further support for the importance of the current study,
Smollar (1999) emphasizes that:
The prospects of homeless youth making a successful transition
from childhood to adulthood are not positive... .Homeless youth
need greater access to care facilities or family homes that offer
opportunities for developmentally productive interactions, and
youth services professionals need access to information about
what types of interventions are effective, (p. 55)
The major contributions of this thesis will not only be to assess factors and assets
most important in determining successful versus unsuccessful transitions to
adulthood by homeless youth, but will also be to identify effective advocacy and
best practices strategies for assisting these young people in permanently exiting the
streets. A particular strength of this research project is posited by the young adult
3


subjects themselves regarding opportunities and strategies to aid service providers
and families in transitioning homeless youths into productive mainstream adulthood.
Application of Theory from Public
Policy and Public Management
From the field of public affairs, two significant theoretical frameworks may
be accessed to guide discussion and application of research findings from the current
study. The first, involving public policy design as offered by Schneider and Ingram
(1997), addresses the social construction of target populations in which four
disparate policy targets are treated differently by politicians and policy makers on
the basis on their purported deservedness and degree of political power. As
described by these authors:
The social construction of potential target populations interacts with
the extent of political power to form four different kinds of possible
policy targets: advantaged (who are powerful and positively
constructed); contenders (powerful but negatively constructed as
undeserving or greedy); dependents (positively constructed as good
people but relatively needy or helpless who have little or no political
power); and deviants (who have virtually no political power and are
negatively constructed as undeserving, violent, mean, and so forth).
Policy entrepreneurs such as policy makers, interest group leaders,
political parties, media, scientists, and others anticipate how an issue
needs to be framed so that public policies advantageous to their own
cause will appear to be the only rational response, (p. 102)
See Appendix A for a modified summary of these constructions.
4


Where homeless youths fall in terms of this design is of utmost importance to
program implications for this group. The homeless in general have historically been
viewed as deviant while teens in the general population have been considered
problematic but still dependent. Again as summarized by Schneider and Ingram
(ibid.):
More negative constructions portray them as lazy, undisciplined,
immoral, or as lacking intelligence. The homeless are pitied, but not
viewed as entirely blameless for their plight. Teenagers do not carry
constructions as positive as those of younger children, but not usually
negative, either. Teens are seen as needing guidance to thwart their
irresponsibility, (p. 109)
For homeless adolescents combining these constructions, the tendency toward
viewing them as deviant has been exacerbated by their unorthodox dress and by their
depiction as gang-like, threatening, and even dangerous.
It has only been during the past couple of decades that adult advocates have
helped swing the public perception toward viewing these hapless youths in a more
positive light, that is, as more dependent and thus more deserving of public
assistance. This shift in perception has moved policy makers to sway public policy,
public legislation (such as through the federal McKinney-Vento Act of2001,
mandating access to public education for homeless youth), and, ultimately, funding
for programs on their behalf.
5


The final shift in social construction of the target population of homeless
youth might ultimately allow their movement toward contender status, in which
they stand to gain more political power and allocation of a greater amount of public
funding. The qualifier toward is emphasized because it would not be to the
advantage of these young people to leave behind the benefits gained from their
dependent status or to have them again viewed in a more negative light.
The second theoretical framework, more characteristically grounded in public
management research, considers a popular entrepreneurial-type offshoot of public
management known as the best practice movement. As espoused by such
researchers as Behn (1993), this line of inquiry uses ethnographic/experiential
(Lynn, 1996, p. 158) case analysis research on the work of excellent managers or
practitioners to guide public (or private, for that matter) programs in producing
effective results. Interestingly, while only a very few best practices for homeless
youth services were found after an exhaustive search of government and not-for-
profit service agency web sites, no academic articles or books have been forthcoming
concerning such practices with this population. Perhaps the ethnographic nature of
the current research study can be used as a starting point in bridging the gap between
best practice academic research in public management and actual best practices in
the field for homeless youth services.
The backlash of criticism against what some in academic circles consider the
soft nature of best practice research provides an interesting contrast to the actual
6


methodology, best practice findings, and program implications of the current study.
These include assertions that such research is inspired by jurisdictional rather than
intellectual reasons and is essentially dissociated from social science theory (Lynn,
p. 159), that it is limited to situations managers have faced before and ignore[s]
institutional context (Milward, 1996, pp. 307-308), and that its authority is to be
sustained, not by codified expertise and communicable mastery of abstract ideas, but
by the appeal of membership in an elite subculture or clan whose members are bound
together by shared experiences (Lynn, p. 160). In refutation of these assertions, this
thesis first attempts to intertwine theory from the broader social sciences with the
unique theoretical bases of public policy and public management. Second, the focus
of the current findings is specific to institutions serving homeless adolescents, and
thus new interventions will be fine-tuned to this population. Finally, the life histories
of the research subjects have been skillfully codified (that is, categorized using key
words or phrases) through sophisticated qualitative software, allowing emergence of
key concepts and significant trends in the data concerning this population.
Providing an example from his own study of service integration, Milward
also discusses the disparity in best practice research between the very different
views of what constituted success (1996) in service interventions between managers
and their clients, ultimately framing the question:
Whose view of success do we accept, the wise case manager or the
client and family who must deal with the system? While success is
7


difficult to measure, I would prefer to let those with a problem define
it before allowing the definition to come from professional ideology.
(p. 308)
Since recommendations from my study are primarily offered by the young adult
subjects themselves, incorporation of these ideas into best practice models for
homeless youth services would go a long way in attenuating the latter criticism of
best practice research. This also provides a way for managers to problem solve in
the face of new situations resulting from new interventions.
Examining and Extending Developmental Assets Theory
From the social sciences, an additional theoretical framework underlying the
current research involves the extensive and growing network of theories and research
surmising that there are essential external and internal developmental assets (see
Figure 4.1, p. 92) that all adolescents need in order to grow up healthy, competent,
and caring (Benson et al., 1999; Search Institute, 2000). It might be predicted that
the greater the numbers of these assets present during this critical developmental
period, the greater the ability of youths to transition successfully to independent
living as adults. This study will attempt to identify a number of key assets which,
through their presence or absence, will affect whether homeless adolescents will be
resilient to, or instead succumb to, homelessness into adulthood. The differences in
assets available to housed versus homeless youths will also be compared, as the
8


results of this research are unwavering in their insistence that this availability is
inherently different between tiie groups.
A second concern of this project will focus on another set of factors, or
disabling conditions, faced by runaway and homeless youths that are often not
experienced (at least not to the same degree) by their housed counterparts living in
mainstream communities. The narratives of the young adult research participants
will thus be used to identify negative influences or characteristics (such as mental or
physical illness, substance abuse, physical/sexual/emotional abuse, indigence,
absence of education, time Spent in detention facilities, criminal records, or
placement in foster care or group homes) that precipitate homelessness among at-risk
youths. It should be expected that these damaging operatives must be overcome
before affected young people will be able to take advantage of programs designed to
transition them into successful independent living situations as adults. It is a guiding
premise of this research that additional, more basic levels of assets (e.g.,
physiological well-being and safety needs, as posed by Maslow, 1970) must first be
met for homeless adolescents before the higher-order assets posed by the Search
Institute can be addressed.
A third area of significance to the current project is the importance of
redirecting a somewhat different set of internal and external assets (survival
skills), learned and mastered by young people while living on the streets, into more
productive mainstream activities. Skills that allow young people to successfully
9


navigate episodes of homelessness may become detrimental in allowing them to
function in normal environments, yet these skills may serve as an important base
of strength from which these disenfranchised youths may draw. Adult role models
and service providers must become adept at channeling potentially negative energies
into positive ones.
The results of this study offer some preliminary suggestions for extending the
theoretical assets and needs models available in the literature to make them more
relevant to homeless youth populations. Chapter Four will more thoroughly address
these theory-building implications derived from the research findings.
Contributions to Theory and Practice
Additional contributions to theory, as may be applied to homeless youth
populations, are expected from the research findings beyond expansion of current
models of developmental assets. Singular credence will be given to Schneider and
Ingrams (1997) social constructions of target populations from the unique position
of one disadvantaged population in its struggle to extricate itself from a more
negative and powerless social construction to one seen as more deserving of
attention and funding from public programs and funding sources. Gender theory will
hopefully be enriched for this population in that differences can be demonstrated
between the experiences of female and male participants while on the streets as well
as between the factors influencing their abilities to permanently extricate themselves
10


from homelessness. Literature and theory concerning sexual minorities, especially
for gay youths, will ideally be further informed by the poignant but optimistic stories
of two young gay males in their climb from homelessness to independence.
Recognition will optimally be given to application of Prochaskas Transtheoretical
Model involving stages of change (see Appendix C) to the continuum of success
through which homeless adolescents must pass in their transition into mainstream
adulthood. That these youths may climb to a higher success level, only to fall back
again, should be recognized as an oft-repeated pattern in their gradual ascent toward
functional independence and is demonstrated repeatedly by the life histories of the
study participants.
At least equally as important, the results of this investigation are expected to
have significant policy and program implications for direct service providers to
homeless adolescents, for social services agencies overseeing out-of-home
placements for youths, and for policy makers charged with shaping public policy to
address funding for needed interventions. The current research findings offer
significant additions to the best practices data bases in the public management
sector, as may be applied to interventions for homeless adolescents throughout the
United States and potentially more globally. Somewhat unique to this line of
practice, however, will be the guiding voices/suggestions from the subjects
themselves in strengthening programs for this population. More specifically, this
investigator has sought to identify whether and how the needs of homeless youths in
11


Denver are being met through existing agencies such as Urban Peak and what
additional services are needed to foster independent living and to reduce the risk of
future homelessness. Differentiation of the unique factors affecting eventual
outcomes for homeless female versus homeless male adolescents are expected to
provide support for more appropriate, gender-based policy decisions concerning
service provision to these populations. Finally, at least some cursory attention to the
unique factors affecting homeless gay and lesbian youths are expected to open
further dialogue for meaningful interventions for this group.
Overview of the Thesis
This research seeks to answer a number of questions about the factors
impacting youth homelessness, but the overarching purpose of this research was to
investigate how the condition of being homeless as an adolescent affects the risk of
remaining chronically homeless as an adult. With this primary research question in
mind, the following sets of secondary questions served to guide the research process:
Risk Factors for Homelessness
> What uniquely common factors or combination of factors among at-risk
adolescents shape whether they will become homeless?
12


> What factors or combination of factors shape whether homeless adolescents
will successfully transition to mainstream adulthood or will in fact remain
chronically homeless as adults?
> How do the environmental experiences of adolescent females differ from
those of males in shaping resilience to, or continuation of, homelessness into
adulthood?
> How do the environmental influences of adolescent sexual minorities differ
from those of heterosexual youths in shaping resilience to, or continuation of,
homelessness into adulthood?
Policy Orientation
> Is the Forty Developmental Assets model, posed by the Search Institute to
ensure all youths grow up healthy and competent, equally applicable to
homeless as well as to housed adolescents?
> How should homeless adolescents be socially constructed by advocates in
order to be viewed by policy makers, funding sources, and the public in the
most favorable light?
> What are the most promising best practice strategies or opportunities open
to service providers in improving the chances of successful transitions of
homeless adolescents into productive mainstream adulthood?
13


To accomplish the task of addressing these questions, several steps were
taken, the first of which was consulting the scholarly literature from several fields.
Initial attention was given to consulting the wider body of literature concerning
homelessness in general, highlighting the fact that homeless youth have only
relatively recently (that is, in the past 20 years) been deemed worthy of more age-
specific study. At the same time, this increased attention and worthiness has
allowed these young people to be viewed through a more positive social lens,
providing increasing opportunities for public funding of homeless youth services. A
review of best practices approaches, available through not-for-profit and
governmental agencies, will be featured at the end of the chapter.
The literature review further gives specific attention to several common lines
of inquiry dominating studies on homeless youth. The first involved common
problems characteristic of this population, including histories of abuse and neglect,
physical and mental illnesses, substance abuse, and involvement in delinquent or
criminal activities. Concerning the latter, the importance of criminogenic theory
(McCarthy & Hagan, 1991; Hagan & McCarthy, 1997) in keeping kids on the streets
was given brief attention. The second line of inquiry more specifically addressed
risk factors that tend to predispose teens to becoming homeless, including the
problems mentioned above in combination with family dysfunction, disengagement
from mainstream educational systems, and foster care or group home placements. A
third line of inquiry focused on the more recent application of resiliency theory (or
14


coping strategies) as applied to the ability of homeless youths to survive on the
streets.
Particularly relevant to the current research project, significant consideration
was given to any literature shedding light on the perpetuation of homelessness from
adolescence into adulthood. While only a few predictors of this phenomenon have
been posed by even fewer researchers, the most important seems to be the absence of
developmental experiences and assets that tend to make youths unable to move
forward in integrating into mainstream society. At the time of this review, there was
only one comparison study (with a sample size of 12 participants) found that looked
at successful transitions of runaway youths into adulthood, and there were no
studies involving those who were not successful in these transitions.
Two of the last three lines of inquiry followed in the literature review
included studies involving researcher collaboration with homeless youths and the
effects of gender and sexual orientation on youth homelessness. In each case, the
literature was very sparse. Concerning the former, it seemed to this researcher that
absence of hearing the voices of the homeless youths themselves tended only to
marginalize them further in society. Regarding the latter, while there have been a
few studies on homeless females, even fewer have been specific to adolescents, and
the literature on gay, lesbian, bisexual, and transgender (GLBT) homeless youths
was virtually nonexistent other than a few short articles in GLBT journals. The final
line of inquiry involved stages of transition of homeless youths, again with only two
15


recent (since 2000) publications beginning to look at the fact that the road to exiting
the streets is not a constant one but rather involves many stages and an uneven
progression.
Following a thorough review of the pertinent literature in Chapter Two,
Chapter Three outlines the methodology used to address the research questions.
Urban Peak, a 501(c)(3) not-for-profit serving homeless youths from late teens
through age 21, became the primary access site used in the current study to enlist
formerly homeless (as adolescents) young adults for participation in the study. For
wider representation, additional young adults were accessed from the streets in
central Denver, from other shelters/agencies, or from word of mouth knowledge of
formerly-homeless students on the Auraria college campus in downtown Denver.
Extensive fieldwork followed which included primary data collection of the life
histories of 27 different young adults, through a total of 48 personal qualitative
interviews. At least two 60 to 90-minute in-depth, open-ended interviews were
collected from 20 of the subjects, 10 males and 10 females. Further cross-
representation involved five racial/ethnic groups and two gay males.
The secondary sample consisted of participants from two focus groups,
utilized in the research design stages of the project and in aiding validation of
research results. The first involved a Youth Advisory Panel comprised of currently-
homeless youths associated with Urban Peak. The second was a focus group of
16


knowledgeable service providers and university researchers involved with Urban
Peaks Research Committee.
In analyzing the data, a modified grounded theory approach (as guided by
Strauss & Corbin, 1998) was utilized. Following transcription of each of the 48 life
history interviews, analysis was aided through the use of QSR N6, the most recent
version of the NUD*IST software program for qualitative data analysis. This
method allowed for the coding and synthesis of key concepts and trends in the data
that were ultimately incorporated into the research findings. This further led to
identification of representative quotations from the research participants themselves
in illustrating and substantiating frequently-occurring and important ideas.
The research findings begin in Chapter Four, which first provides a more
extensive discussion of the theoretical framework of internal and external
developmental assets needed by all adolescents to become competent adults.
Summarized previously in the current chapter, the assets available to housed
versus homeless youths are more completely reviewed in Chapter Four, and a new
extended model is offered for the latter incorporating a set of basic needs that must
be met for homeless youths before they may be able to access higher-order assets.
Chapter Five presents significant findings concerning the debilitating
conditions that frequently pull street youths in a downward spiral, often drawing
them back and preventing them from permanently exiting the streets. The purpose of
this chapter is to use the voices of the formerly-homeless young adults to shed light
17


on which of these conditions seem most damaging to homeless youths in this
process. While some of these conditions (for example, substance abuse or
irresponsible interactions with the law) tend to cast homeless youths into the role of
negatively-constructed deviants, as described by Schneider and Ingram, other
conditions (e.g., physical or sexual abuse in their homes of origin) allow these youths
to be viewed more positively as dependents, weak and helpless but worthy of care.
Of particular interest here are two results-based models, one of which depicts
vicious cycles keeping homeless youths entrapped in behavioral sequences that
prevent them from transitioning into productive adulthood. The second presents a
more linear diagram illustrating factors leading to initial adolescent homelessness,
with such homelessness in turn prompting debilitating factors keeping youths on the
streets.
The last chapter of research findings, Chapter Six, alternatively provides a
significant and revealing overview of those internal and external assets most likely to
aid homeless youths in their quests to permanently extricate themselves from the
streets. From the narratives of a number of the young adult subjects come stories of
amazing internal resilience to the crippling conditions that keep less resilient
youths in a chronic pattern of homelessness. These internal assets allow them to put
behind the negative episodes in their lives and move forward with ever-stronger
determination. Moreover, these young people describe the external assets provided
18


by service agencies, programs, interventions, and individuals, all of which were
considered instrumental in helping them up and out of chronic homelessness.
Chapter Seven provides amazingly perceptive and far-sighted
recommendations from the formerly-homeless young adult participants concerning
strategies and opportunities for homeless youth services providers, public policy
makers, public institutions (schools, social services, the justice system), and
families/guardians to improve the chances of successful transitions of todays
homeless youths into tomorrows productive adults. Additionally, advice directed
toward the youths themselves includes no-nonsense admonishments to quit
whining and help themselves out of their current situations. Extension of best
practice research in public management to include many of their recommendations
will not only serve to legitimize best practices for homeless youth services but will
also lend support to the importance of collaborative research with target populations.
Finally, Chapter Eight provides a summary of conclusions and implications
for policy and further research. The initial overview of significant findings
summarizes the most important factors shaping adolescent homelessness from the
perspective of study participants as well as those factors affecting adult pathways
into/out of homelessness. It further brings together principal differences in findings
by gender and sexual orientation. The next section, of primary significance to this
study, discusses policy and program implications, demonstrating how current
research findings may be used to improve effective advocacy for homeless youths, to
19


develop best practice models for youth services agencies, and to use positive youth
assets in helping them exit the streets. This includes the importance for public policy
of constructing this population in a more positive light, worthy of the maximum
amount of support and public funding that society can allow in order to move them
into productive mainstream living. Extension of the Search Institutes Forty
Developmental Assets Model to better accommodate homeless young people is
reiterated as another distinguishing contribution of this thesis. The final section of
this chapter is directed toward posing recommendations for future research efforts
related to the studys principal research questions and major themes.
20


CHAPTER 2
REVIEW OF THE LITERATURE
While attention to homelessness in general has become of increasing social
concern during the past few decades, most of the literature of die 1970s and 1980s
addressed issues of mental health, physical diseases, substance abuse, housing
concerns, unemployment, and social welfare and health policy issues related to
homeless adults. Not until the late 1980s and early 1990s did issues concerning
homeless children and families begin to surface, and there remain amazing gaps in
rigorous research on homelessness among adolescent youths in the United States.
This chapter will review the existing academic literature concerning common
lines of inquiry dominating studies on homeless youth. The first three themes, more
commonly addressed by prior researchers, involve common problems of this
population, risk factors that predispose teens to becoming homeless, and application
of resiliency theory to this population. The last four lines of inquiry, for which there
is very littie in the existing literature but which are particularly relevant to this thesis,
involve predictors of homeless adolescents remaining homeless into adulthood,
researcher collaboration with homeless youths, the effects of gender and sexual
orientation on youth homelessness, and the uneven stages of transition out of
homelessness for this population. Finally, while not present in the academic
21


literature of public policy and public management, some attention will be given to
best practice interventions for these young people as viewed by service providers
and to best practice models in the literature of related academic areas (such as found
in the child abuse, criminal justice, and drug abuse literature).
The ultimate goals of this review are to inform the reader not only of gaps in
the literature on youth homelessness, but also to lay the groundwork for expanding
what is known in public policy and public management research by utilizing insights
from the lived experiences of formerly homeless youths. Mayers (2001) captures the
justifications for this quest:
[T]he literature has not provided a comprehensive picture of the
experiences of these youngsters. For the most part, the literature
is fragmented and compartmentalized. Issue by issue, researchers
have dissected and investigated components of the phenomenon,
leaving trails of the parts without any sense of the whole. In an
effort to explicate certain aspects of the street, the literature has
become de-humanized. (p. 160)
Expanding Attention to Adolescent Homelessness
Even into the 1990s the majority of attention given to homeless youths has
been relegated to sections or chapters of books on general homeless populations
(Brickner et al., 1990; Robertson & Greenblatt, 1992; Fisher & Collins, 1993), on
more generalized ages of homeless children (Kryder-Coe et al., 1991; Shane, 1996;
Sullivan, 1997), or in the context of strengthening services to homeless and runaway
22


children and youths through working with families (Rothman, 1991; Bass, 1992).
Still another group of studies (Russell, 1995, and the majority of professional journal
articles) has dealt with one specific aspect or risk factor associated with
homelessness across a variety of ages, including child maltreatment, mental
disorders, substance abuse, sexually-transmitted diseases, survival sex (sex in
exchange for food, money, or shelter), suicidal behavior, hepatitis B, or juvenile
delinquency. Liddiard and Hutson (1991) describe the limited options for housing
available to homeless youth, noting that some options (returning home or staying
with friends) become less available over time, while squatting in abandoned
buildings or sleeping rough (sleeping outside) become more frequent options the
longer young people remain homeless.
It has not been until recent years that dealing with the growing population of
homeless youths on their own has been given primary attention along multi-
dimensional lines. The relatively small volume of literature that does exist to date
focuses on the maladaptive behaviors of these young people but rarely on their
potential for positive solutions to overcoming these barriers. Van der Ploeg (1997)
and Whitbeck and Hoyt (1999), although addressing relevant aspects of the
dysfunctional family lives of homeless youths and parent/caretaker adolescent
relationships, focus their primary attention on the characteristics and problems of
adolescents literally on their own. In this context, Mayers hypothesizes, Living on
the street reconstructs to some degree the abusive relationships adolescents
23


experienced at home. The reality of the street includes being hungry, lonely, scared,
cold, and a willingness to do whatever it takes to survive (2001, p. 153). Tienda
and Wilson (2002, p. 8) lament the marginalized status of such youth that ignores
or undervalues them because much of the current literature on youth focuses on the
assessment of the negative consequences of maladaptive behavior ... rather than on
the potential of youth for creative change.
Common Problems Characteristic of Homeless Youth
For purposes of the current study, several lines of inquiry have dominated
studies on homeless youths. The most frequent involve common problems
characteristic of homeless youth. Kryder-Coe et al. (1991) and Robertson and
Greenblatt (1992) enumerate a wide array of mental and physical health problems
experienced by this population, with the former emphasizing the importance of the
homeless child syndrome, comprised of poverty related health problems,
immunization delays, untreated or under-treated acute and chronic illnesses,
unrecognized disorders, school, behavioral problems, child abuse, and neglect
(1991, p. 96). Brickner et al. (1990, p. 9) relate that trauma, including that
associated with rape, sexual abuse, suicide, and attempted suicide, is frighteningly
common among homeless youth. Physical and sexual victimization (Tyler et al.,
2001; Whitbeck et al., 2001), depression and suicidal behavior (Molnar et al., 1998;
Rohde et al., 2001), and violence and youth crime (Hagan & McCarthy, 1997;
24


Wardhaugh, 2000; Baron et al., 2001) have been particularly prominent in the
literature concerning common characteristics of homeless youth. For example,
McCarthy and Hagan conclude that street life is inherently criminogenic, lending
itself to the necessity of committing criminal acts in order to survive, and that the
longer the homeless episode, the greater the involvement in crime (1991, p. 408).
Nationally, the percentage of homeless and runaway adolescents reported to be in
trouble with the juvenile justice system alone is 30 percent (Schaffher, 1999, p. 25).
Unlike chronically homeless adults, however, who were more likely to have been
incarcerated prior to becoming homeless, homeless youths were more likely to have
experienced jail time or incarceration for their involvement in crime after becoming
homeless as a response to demands of being on the street (Hagan & McCarthy,
1997, p. 9).
Similarly, a number of studies have addressed the prevalence of acute and
chronic illnesses in this population. A study by Beech et al. (2002) found 22% of the
homeless adolescents they sampled (N = 150) tested positive for hepatitis B or C,
related to the high risk behaviors many of them engage in (that is, sharing
contaminated needles or exchanging sex for food, shelter, or drugs). The relatively
high prevalence of sexually-transmitted diseases, such as Chlamydia or herpes or
AIDS, is also the focus of numerous articles concerning homeless or runaway youth
(Rew, 1996; Rohde et al., 2001; Van Leeuwen et al., 2002). Fisher and Collins
(1993) stress the major characteristics of alcohol and drug abuse, while McMorris et
25


al. (2002, p. 39) go as far as to equate such alcohol abuse with a vicious cycle in
which homeless youth go on to become chemically dependent and chronically
homeless adults.
Risk Factors that Predispose Youths to Becoming Homeless
A second common line of inquiry has concerned specific risk factors that
predispose youths to becoming homeless in the first place. Kryder-Coe et al. (1991,
p. 52) summarize that groups at risk for homelessness appear to be young people in
such systems of care as foster care, group homes, juvenile detention, and other
institutional settings. Another group at risk may be youth with serious problems in
their schooling, special education needs, or mental health problems. These authors
further note, Parent alcoholism appears to be a significant but hidden factor in
homelessness for many adolescents (1991, p. 42). Schaffiier asserts, Teenagers
often consider running away as the best solution to family struggles, especially when
chronic and acute physical and/or sexual abuse by parents has occurred (1999, p.
64). Brickner et al. (1990, p. 9) take the stance that the largest estimates include not
only children who run away from home but also those who are kicked out
(throwaways) and those whose homes evaporated because they grew too old for
foster care placements. Homeless youths, in general, are alienated from their
families and lack education and job skills, giving them nowhere to go from there.
A particularly poignant article in the Los Angeles Daily Journal (Romo, 2001, p. 1)
26


addresses the issue of how emancipation from foster care leaves many young people
literally on the streets. More specifically, Cauce et al. (2000, p. 236) relate that a
third of all homeless and runaway youth in their study had prior foster home
placements.
Robertson and Greenblatt (1992, p. 291) reiterate many of these same
themes, stating that half of these youths.... had been seriously abused physically,
had been actively rejected, or were agency youths who depended on public
institutions for food and shelter. For these adolescents there was no safe home to
which they could return. In this same book Petry and Avent (1992, p. 299)
conclude that although there are numerous reasons why youths leave home, one of
the most clearly defined is the breakdown of family ties at the core of which is
deterioration of communication between the youth and the adult in his or her life.
Schaflher asserts that running away consists of emotional and moral crises of trust
(1999, p. 45). James (1991, p. 22) supports these theories in reporting the results of a
1991-92 survey of over 200 homeless Colorado youths on their own who were in
contact with service agencies. While the youth believed the root problem to be
family or parental failure, four main events were identified as precipitating literal
homelessness among adolescents: running away from home, being told by parents to
leave, abandonment, or being removed from their homes by social services or other
government agencies. To summarize the findings concerning factors that predispose
27


youths to becoming homeless in the first place, perhaps Downer (2001) says it most
succinctly:
There is no single pathway into homelessness. The myriad of needs
that the homeless population presents with reflect the variety of
experiences that propel them into homelessness, (p. 4)
Further, Mayers describes the complexity of this issue by posing a question in return:
I am tempted to ask why some abused kids are inspired to leave home while others
are not (2001, p. 147).
Resiliency in Homeless Adolescents
A third and more recent line of inquiry related to the current study has been
application of research on resiliency (i.e., coping mechanisms) to the adverse effects
experienced by homeless youths. While such authors as Luthar (1991) and Schaflher
(1999) have addressed resiliency in high-risk adolescents in general (including both
successful and unsuccessful strategies), it wasnt until an extensive study by Rew et
al. (2001) that specific consideration was given to resilience and connectedness as
protective factors in homeless adolescents, making those who perceived themselves
as resilient less hopeless, more connected, less engaged in life-threatening behaviors,
more adaptive to street life, but also overly self-reliant. While Mayers (2001, p. 149)
acknowledges youth who engage in criminal behavior, even if it is petty theft, are
rarely categorized as being resilient, malleable, adaptable to the life world of the
28


street, she contends that such youth do find a way to survive, pointing to the need
to understand this phenomenon in terms of positive survival skills. Seeming to
answer this challenge, Rew and Homer (2003), emphasizing the need to focus on
strengths (rather than just limitations) of homeless adolescents that contribute to their
health and well-being, have identified both resources, serving as the foundation for
survival, and self-improvement actions that assist youths in exiting the streets and
becoming productive members of society. In summary, Wardhaugh (2000) provides
a realistic but hopeful note concerning the role of resiliency:
Amid social and political developments, young homeless people
continue to lead their lives, some caught up in the daily round of
subsistence and others looking ahead with hope and apprehension -
to the future, (p. 132)
On Homeless Youths Becoming Homeless Adults
A related line of inquiry involves concern for whether homeless youths will
become the next generation of homeless adults and possible supposition regarding
predictive factors associated with this risk. Drawing from attachment theory,
Mayers suggests, The ramifications of having poor attachment [to parents or other
adults] may play a role in why some kids maintain their lives on the street (2001, p.
146). Kryder-Coe et al. (1991) and James (1999) point to disproportionate numbers
of homeless adults who were homeless as children who have reported out-of-home
placements (in foster homes, group homes, detention centers, and other institutional
29


placements) as minors. Downer (2001, p. 5) reports that teen parents not only make
up a significant number (20-50%) of homeless families seeking shelter but that also
many of these parents were graduates of the child welfare system. In a special
report of Homes for the Homeless (1997) the connection between foster care and
future adult homelessness has been explored in depth. One alarming example
involved a survey of nearly 400 homeless parents in New York City in which 20%
had lived in foster care as children and 20% currently had one or more children in
foster care. Woodhouse (2002, p. 10) goes on to relate that youths in foster care or
other agency placements graduate to adulthood with no human capital and no safety
net and that between 25% and 35% had been homeless for periods of time or
remained homeless after exiting foster home placements.
Robertson and Greenblatt (1992) discuss two other powerful predictors of
maintenance of homelessness into adulthood. The first concerns low levels of
current family contact and support (an asset), while the second involves a process
of adaptation to homelessness in which a change in self-perception and self-identity
both facilitates the individuals current adjustment (to a homeless status) and
contributes to the maintenance of homelessness (1992, p. 13). Often the very
behaviors that are necessary for youths to survive on the streets particularly crime,
physical violence, and drug possession lead to felony records that make later
employment difficult. Hagan and McCarthy (1997, p. 204) suggest that
unemployed and criminally involved street youth are unlikely to experience a
30


successful transition to adulthood. Whitbeck and Hoyt (1999) speak not only to the
latter concerns, but also to the absence of key developmental assets which serves as
an undergirding theoretical framework for the current research study:
Chronic runaways and homeless adolescents are literally learning to
become marginal adults. For many runaway and homeless adolescents,
early adult status has come at the cost of essential developmental
experiences. These developmental losses are difficult to remedy
under the best of circumstances and currently the resources available
to intervene are woefully inadequate. By not responding to the needs
of runaway and homeless adolescents we are choosing to train them to
become marginal adults, part of the criminal justice system, perhaps
even part of the adult culture of homelessness, (p. 171)
The Importance of Developmental Assets in
Preventing Future Homelessness
Concerning the importance of developmental assets in producing competent
and independent adults, Benson et al. (1999) and the Search Institute (2000) provide
a framework of essential internal and external assets that all young people for
successful transitions to adulthood (see Figure 4.1, p. 92). The rationale for and
implementation of their 40-asset model is discussed in some detail in the
introductory chapter (see Chapter 1, pp. 8-10). Included in the list of external assets
are support systems provided by significant individuals and institutions in a youths
environment, such as involvement and support of parents, adult and peer mentors,
schools, religious organizations, and the community at large. The list of internal
assets includes self-generated strengths and value systems that protect young people
31


from negative environmental influences and propel them toward self-actualization
and healthy independence.
Other authors, including developmental psychologists, further underscore that
successful normative youth development requires intact nuclear and extended
families, supportive communities, realistic perceptions of opportunities, and
predictable behavior and experiences (Tienda & Wilson, 2002, p. 9). These authors
further argue that only confident youth who dare to imagine a better future are able
to focus their energy on reaching it (ibid.). Schaffner observes that children need
love and guidance in supportive environments to make the transition from
adolescence to healthy adulthood (1999, p. 141).
Yet nowhere in the literature is there more than lip service given as to how
homeless youths, lacking in even basic subsistence assets (food, shelter, adequate
clothing, medical care), much less the afore-described assets necessary for proper
development, can successfully transcend the all-encompassing energy sinks of
basic survival to become confident and competent adults. An important contribution
of the current study will be to expand the Forty Developmental Assets model of the
Search Institute to include a more basic set of needs that must be met for homeless
adolescents before higher levels may be addressed. Inclusion of the lowest rungs of
Abraham Maslows Hierarchy of Needs (Maslow, 1970), especially Physiological
Needs and Safety Needs, should serve as a point of departure toward that end (see
Appendix B).
32


Other Factors and Predictors
A few comparative studies have specifically compared homeless youths with
youths in the general population (that is, those with fewer versus greater numbers of
developmental assets). Greene and Ringwalt (1998) compared the prevalence of
pregnancy among runaway and homeless youths with that in the general youth
population. Jones (1998) compared health status and service use of adolescents at a
school-based health clinic with homeless adolescents. Wolfe et al. (1999) did a
cross-comparison study of homeless and matched housed adolescents on family
environment variables. In all of these studies the homeless youths fell short of their
housed peers. Downer (2001, p. 11) summarized a number of additional studies
comparing homeless and domiciled adolescents, revealing that homeless youths are
significantly more deprived emotionally, socially, and culturally and tend to report
lower levels of parental care and acceptance. Concerning comparisons of school
attendance between housed and homeless students, Rafferty and Rollins (1989)
reported poor attendance levels for homeless children at all levels in New York City
public schools, with attendance rates at 63.6% (vs. 85.5% for all students) at the
junior high level, and 50.9% (vs. 83.9%) at the high school level. On the other hand,
Rohde et al. (1999) compared the IQ scores among homeless older adolescents with
their housed counterparts and found average intellectual performance to be
essentially comparable.
33


Few if any comparison studies have been done on predictions of whether
homeless youths will become homeless adults or will successfully transition to
independent living in adulthood. Lindsey et al. (2000) recently published a
qualitative study addressing the personal strengths and resources (e.g., attitudes and
spirituality) that enabled 12 formerly runaway or homeless youths in North Carolina
and Georgia to make successful transitions to adulthood. However, this study did
not access young adults who were not successful in these transitions (e.g., were still
homeless), did not provide for a continuum of change or success, and depended
entirely upon a self-defined sense of success from participants, ignoring societal
definitions.
Other authors have emphasized those factors that seem to preclude homeless
adolescents from exiting the streets and from successfully integrating into society.
Tienda and Wilson note that [r]ising residential segregation, which often acts as a
mobility trap, produces social isolation, resignation from required and optional
activities, rejection of mainstream norms, unemployment, and violence (2002, p.
11). Obviously, homeless youths epitomize disadvantaged groups socially isolated
from mainstream social institutions (nuclear families, schools, places of worship),
encouraging them to develop maladaptive strategies as they negotiate the
developmental challenges of adolescence (ibid., p. 13).
34


Researcher Collaboration with Homeless Youths
While such studies remain few, a number of relatively recent publications
involving researcher collaboration with homeless youths provide still another line of
common inquiry concerning this population. In the first known study, published as
her dissertation, Gay (1996) explored the effect of homelessness as related to the
experiences of a group of 18 and 19 year olds from two Chicago shelters. Pfeffer
(1997) sought to provide a forum for young homeless and runaway women to voice
the unique circumstances of their life experiences. Ensign and Gittelsohn (1998)
investigated the perceptions of the health needs of shelter-based homeless youths in
Baltimore, Maryland. Still another article addressed forming collaborations with
suburban street youths to explore HIV-related risk rates and prevention strategies
(Harper & Carver, 1999). Klaus J. Jacobs (Tienda & Wilson, 2002, p. xii)
emphasizes the conviction that the views of young people must be integral to the
formulation of policies aimed at improving their life chances. In this same volume,
Felton Earls and Mary Carlson devoted an entire chapter to Adolescents as
Collaborators: In Search of Well-Being (ibid., pp. 58-83). The case for youth-
driven policy, including development of policy priorities and involvement of teens in
the policymaking process, is summarized in an article distributed by the California
Center for Civic Participation, titled Engaging Youth in Policymaking Improves
Policies and Youth Outcomes (Balas, 2003, p. 1).
35


No articles have been found, however, involving collaboration with young
adults who have successfully transitioned out of homelessness to become housed
adults (versus those who have not) after episodes of homelessness as youths. This
study breaks new ground through researcher collaboration with young adults who
have succeeded (and some who have not) in transitions out of homelessness:
Effects of Gender and Sexual Orientation
on Youth Homelessness
There is emerging evidence of gender differences among homeless youths,
long-ignored in the literature. Studies exploring the unique characteristics of
homeless youths with differing sexual orientation are even less common.
Concerning the former, Pfeifer tells us that historically.... girls are mis-categorized
and mis-represented because researchers use male models of life experiences and
gender-based morality as the framework for the studies (1997, p. 33). Whitbeck
and Hoyt note, without regard to shelter arrangements, adolescent males and
females tend to run away at about the same rates (1999, p. 7).
To be sure, some studies (Welsh et al., 1995; Hagan & McCarthy, 1997)
mention a few specific differences between female and male runaways or homeless
adolescents, such as lower self-esteem and higher frequency of physical and sexual
abuse at home for females. Cauce et al. (2000, p. 236) more pointedly relate that
60% of girls in their study reported sexual abuse prior to leaving home as opposed to
36


23% of boys, although boys were more often physically abused, assaulted, or
arrested. More specifically, while 80% of youths who have been sexually abused
before leaving home are female (Welsh et al., 1995, p. 29), approximately 97% of
the perpetrators are males known to their victims, with 40% to over 50% functioning
as a parent (father or step-father) at the time of the abuse (Welsh, ibid.; Hatty et al.,
1996, p. 170).
Passaro addresses homeless women in general, relating that homelessness
seems like their last chance at survival, because it affords immediate escape as well
as what they hope will be the opportunity to recreate homes without danger that is,
homes without men (1996, p. 72). She goes on to cite other differences from their
male counterparts, i.e., they are not as confrontational, they are less visible, and they
are more likely to receive spare change from passersby (ibid., p. 87). Sheets (1999,
p. 27) discusses common threads running through the stories of homeless women,
including not having taken sufficient steps to ensure their financial or physical
security, repeatedly choosing bad companions, and hanging on to an old paradigm
or behavior pattern that cast themselves in the role of pitiful victims.
Concerning homeless female adolescents, however, Pfeffer (1997) laments
that:
There are no studies to date which focus on young women,
investigating the circumstances of their life experiences. However,
the youth homelessness of the 1990s, especially among girls, is a
complex arrangement negotiated by urban developers, advocates,
37


the street economy, youth subcultures and the moralistic
representation and relative invisibility of girls who live on-their-
own. (p. 32)
Interestingly, however, Fitzpatrick did begin to shed some light on the gender
differences concerning young homeless women, reaching the conclusions that they
are far less visibly homeless than young males (especially before the age of 18) and
that they are more willing than young men to approach formal agencies when they
find themselves homeless, and thus are less likely to take hidden routes through
homelessness (2000, p. 78). She also offered that the young women in her study
seemed to move out of homeless situations more successfully than young men
(ibid., p. 76).
Even more apparent in the social science literature, through the end of the
twentieth century, has been the near absence of studies concerning the specific
challenges faced by homeless sexual minorities (youths with gay, lesbian,
bisexual, or transgender orientations). A preliminary 1995 study (Buchanan, 1995,
p. 42) merely reported the presence of increasing numbers of gay teenagers on the
streets of New York, San Francisco, and Portland, with many of them having been
pushed out by their families who cant cope with their sexual orientation. Another
study, focusing primarily on female delinquency (Schaffher, 1998, p. 8), did provide
estimates that the proportion of gay and lesbian runaway and homeless youth to be
as high as 40% of the total street youth population. Cauce et al. (2000, p. 235),
38


however, identified only 7% of their subjects as non-heterosexual. At any rate,
virtually all researchers addressing this population agree that they are at even higher
risk for adverse experiences than their heterosexual counterparts. For instance,
Cochran et al. (2002) found that homeless adolescents with GLBT membership are at
increased risk for negative outcomes while on the streets, including more frequent
victimization, psychopathology, and substance abuse addiction. Mayers (2001, p.
159) describes increased levels of stress and increased risk of suicide among gay and
lesbian homeless adolescents.
Stages of Transition of Homeless Youth
In anticipating possible objections or caution flags which might arise by
readers of this study concerning how one might reasonably be sure into which
category the young adult subjects should fit, particularly when there would appear to
be shades or a continuum of success experienced by formerly homeless youths
attempting to transition out of homelessness, this researcher again has turned to the
literature to better inform and respond to such objections. One such model, dubbed
the Transtheoretical Model by its originator, James Prochaska (Prochaska et al.,
1990, p. 60), uses stages of change to integrate processes and principles of change
from across major theories of intervention (see Appendix C). In the spirit of
recognition that study participants may indeed fall along a continuum of success in
their transition into independent adulthood and out of homelessness, I attempted to
39


structure some of the questions in the open-ended interviews to ascertain into which
stage along the continuum each participant might fit.
Fitzpatrick more specifically decries the fact that most researchers imply one
main pathway through which most homeless youth must navigate, with some
managing to exit homelessness at various stages while the rest descend into
chronic homelessness (2000, p. 73). She goes on to describe a range of six
different pathways that the youths in her study took through homelessness, with
clear distinctions in the progress of young people who followed different pathways
(ibid., p. 75). Of particular interest, those who became associated with adult shelters
and those who were visible in the city centre (such as youths visible on the 16th
Street Mall in downtown Denver) were the two pathways taken by youths most
likely to become entrapped in the downward spiral toward chronic homelessness.
On the other hand, those staying in accommodations described as the local official
network (that is, youth shelters) made the best progress toward exiting
homelessness. Their employment and housing circumstances were much better
than the other groups, Fitzpatrick explains, and they had the closest contact with
their families (ibid., p, 139).
Another more recent model, using a life cycle approach to youth
homelessness, has been proposed by Auerswald and Eyre (2002), who also offer a
number of stages through which street youths typically pass in their sojourns on the
streets. According to these authors, The life cycle model suggests that street youth
40


who are most open to intervention are those who are in transitional states, i.e., those
who have just arrived on the street or those who are in crisis (disequilibrium) (ibid.,
p. 1497). These young people thus have the greatest probability of finding a new
identity in mainstream society (ibid.) and thus transitioning into successful
adulthood. Those unable to extricate themselves from street life may invariably
return to the streets in repeated episodes of recidivism, placing themselves at highest
risk for chronic homelessness as adults.
Best Practices in Interventions for Homeless Adolescents
Chapter One initiated discussion on the entrepreneurial-type offshoot of
public management known as the best practice movement. Overman and Boyd
(1994, p. 68) describe best practices research (BPR) as a method of inductive
practice-to-principles research of the postbureaucratric genre that is (using their
descriptors) pragmatic, practice driven, innovative and entrepreneurial, positive and
prescriptive, and commercial and user friendly. They go on to critique its practical
and scientific challenges (ibid., pp. 76-80), including that BPR:
Does not learn from experience.
Does not listen to all practitioners.
Does not accumulate practice wisdom.
Is not transferable.
Is not theory-testing research.
Violates threats to validity.
Is not sufficiently probing or critical.
41


That being said, practitioners and university researchers have continued to develop
and refine various versions of BPR, touting such tools as benchmarking (Keehley
et al., 1997) and innovative problem solving coupled with theory (Holzer &
Callahan, 1998) to improve best practices in the public sector. A contribution of the
current research is aimed at addressing some of the afore-mentioned issues,
especially concerning use of collaborative research with research participants and
service providers (through their collective experiences) and accessing theory in
public policy and public management to guide development of the research design.
As noted at the beginning of this chapter, a review of the literature revealed
no academic articles or books concerning best practices in public management
with the homeless youth population. However, an exhaustive search of government
and not-for-profit service agency web sites did impart a few noteworthy best practice
sites for homeless youth service programs. The largest site, from the National
Alliance to End Homelessness, recognized 46 program models across the nation for
their dedication to preventing and ending homelessness (internet citation). Only six
of these addressed interventions for homeless adolescents, with the first three being
nonprofit and the last three government-sponsored:
Project SAFE, in Everett, Washington, is an emergency prevention program
for parents and caretakers at risk for homelessness.
Lighthouse Youth Services, in Cincinnati, Ohio, provides housing options for
at-risk youths and young adults, including an Independent Living Program
that targets foster youth and juvenile offenders, and a Transitional Living
Program targeting homeless youth 18-25.
42


The Larkin Street Youth Services (LSYS) HIRE UP Program in San
Francisco, California, is a workforce development initiative offering
educational and employment services in conjunction with housing and case
management support.
The Adolescent Transition Groups, sponsored by the State of New Mexico,
supports the transition of vulnerable youth from out-of-home care into
independent living.
The Youth Housing Assistance Program, established by the Illinois
Department of Children and Families, provides housing advocacy and cash
assistance for youths 18-21 emancipated from the foster care system and at
risk for homelessness.
The Connecticut Department of Children and Families Housing Continuum
provides a continuum of housing options for emancipated youths 14 and
older.
Two other best practices sites for homeless youths are also worthy of
mention as they relate to this study. Multnomah Countys Homeless Youth
Continuum of Care, provided by the Office of School and Community Partnerships
in Portland, Oregon, offers homeless youths 13-21 a broad range of shelter, outreach,
housing, health care, education, case management, counseling, and employment
services, very similar to those provided by Urban Peak in Denver.
The Colorado Office of Homeless Youth Services has been recognized for its
role in coordinating the efforts of public and private agencies serving Colorados
homeless youths between 15 and 21 years of age. Implementation of the federal
McKinney-Vento Homeless Education Act of2001 (Title X, Part C, of the No Child
Left Behind Act), enacted to ensure access to public education and success in school,
has received primary emphasis.
43


In addition to best practice models developed by service providers for
homeless youths, current and future research aimed toward best practices in public
management of these youths may draw upon BPR in related areas of inquiry as they
apply to this population. For example, such models are already present in the
literature for child abuse (Kelly, 2001), drug and alcohol abuse (Gorman, 2003), and
juvenile justice (Blomberg & Waldo, 2002). Using the methodology and theory
derived from these policy arenas, clearly much farther along in best practices
treatment models, will allow researchers in BPR for homeless youths a significant
jump-start in developing similar models for effective interventions for this
population.
Summary of Literature Review
In summary, while some aspects of adolescent homelessness have been
researched more extensively than others, there are significant gaps in the literature
that should be important to policy makers in formulating relevant public policy and
to service providers in providing effective interventions for this constituency. The
results of the current study, presented in Chapters Four though Seven, fill some of
these gaps, particularly regarding the effects of gender and sexual orientation on
homelessness, the importance of researcher collaboration with the informants, the
uneven transition out of homelessness, and the importance of an extended model of
44


assets needed by homeless youths to transition out of homelessness and become
competent adults.
Most importantly, gaps in the academic literature for effective policy
advocacy on behalf of homeless youths and for best practices in public management
of this group served to guide development of both the research questions and the
research design of this project. The next chapter addresses the methodology utilized
in data collection and analysis and in synthesis of research findings.
45


CHAPTER 3
DESIGN OF THE RESEARCH
The purpose of this research was to compare and analyze the perceptions of
young adults, all of whom were formerly homeless as adolescents, regarding the
factors or influences shaping whether they were able to successfully transition to
independent living in adulthood, whether they remained chronically homeless as
adults, or whether they were currently in transition in their quest for productive lives
beyond the streets. In so doing, I first pose the central research questions, initially
stated in Chapter One but reiterated here because they served to guide the research
design. After discussing the rationale for the use of qualitative research
methodology in addressing the principal research questions, the remainder of the
current chapter will address site of the current study, primary and secondary sample
criteria and selection, data collection including in-depth interviewing techniques,
analysis of the data, and issues concerning credibility, ethics, and limitations.
Special attention will be given to the selection and application of a data analysis
computer software package, QSR NUD*IST Version N6, to improve the uniformity
and efficiency of the analysis.
The following principal research questions, divided into two categories by
content type, were posed to guide the research design:
46


Risk Factors for Homelessness
What uniquely common factors or combination of factors among at-risk
adolescents shape whether they will become homeless?
What factors or combination of factors shape whether homeless adolescents
will successfully transition to mainstream adulthood or will in fact remain
chronically homeless as adults?
How do the environmental experiences of adolescent females differ from
those of males in shaping resilience to, or continuation of, homelessness into
adulthood?
How do the environmental influences of adolescent sexual minorities differ
from those of heterosexual youths in shaping resilience to, or continuation of,
homelessness into adulthood?
Policy Orientation
Is the Forty Developmental Assets model, posed by the Search Institute to
ensure all youths grow up healthy and competent, equally applicable to
homeless as well as to housed adolescents?
How should homeless adolescents be socially constructed by advocates in
order to be viewed by policy makers, funding sources, and the public in the
most favorable light?
What are the most promising best practice strategies or opportunities open
to service providers in improving the chances of successful transitions of
homeless adolescents into productive mainstream adulthood?
The first set of questions, under the framework Risk Factors for
Homelessness, involved the identification of common factors serving either as
precursors to initial homelessness or as disabling conditions that kept these young
people chronically on the streets. Comparisons were also made by gender and sexual
orientation. The second set of questions, falling under the rubric of Policy
47


Orientation, was intended to lay the groundwork for better policy advocacy and
improved service provision on behalf of homeless adolescents. Thus, these questions
were specifically oriented toward three policy goals:
To assist advocates for homeless youths in swaying policy makers and
funding sources to view these youths in a more positive light, deserving of
attention and monetary support.
To develop a hierarchy of developmental assets more suitable to this
population that would include both basic (physiological and safety) needs and
the unique survival skills of these youths, in order that public policies
would support these lower-level assets for marginalized teens.
To promote best practice models for service providers to better facilitate
successful transitions of their young clients to healthy and productive
adulthood.
On the Importance of Qualitative Research and Life Histories
Use of qualitative research methodology, directed toward building grounded
theory (Strauss & Corbin, 1998) and consisting of in-depth retrospective interviews
of formerly homeless young adults, seemed to be the most reasonable approach for
answering the research questions posed for this study. As researchers seek to find
the patterns in the experiences of homeless teens and young adults, it is important to
go back to the beginning to hear their life stories and to try to understand. In the
words of Marshall and Rossman (1999):
For a study focusing on individuals lived experience, the researcher
could argue that one cannot understand human actions without
48


understanding the meaning that participants attribute to those
actions their thoughts, feelings, beliefs, values, and assumptive
worlds; the researcher, therefore, needs to understand the deeper
perspectives captured through face-to-face interaction, (p. 57)
Locke et al. (2000, p. 116), in supporting the use of qualitative research designs,
stated, Here is a form of research that invites questions that deal with how real
people think and feel. This approach is further appropriate for exploratory research
and is in keeping with Robertson and Toros (1999, pp. 3-21) recommendation that
qualitative interviewing techniques (especially of the subjects themselves) be used in
assessing the needs of homeless youths.
Perhaps more compelling in support of using qualitative over, or at least in
addition to, quantitative methods in future research with this population, Mayers
decries the staggering gaps in the literature in providing a comprehensive view of
the lived experiences of these young people, leaving trails of the parts without any
sense of the whole (2001, p. 160), perhaps because the research methods most
often used have no place for stories (ibid., p. 136). More pointedly, she implicates
quantitative academic research for its failure to take a holistic approach in
addressing the complexities of street life:
There has been a focus on overt, behavioral, and measurable
aspects of homelessness and somewhat of a disregard for the
experiences therein, that is, the lived realities of youth have
been reduced to prevalence rates, (p. 160)
49


Of particular interest, then, for this retrospective study, was the importance of
taking life histories of participants during the in-depth interviews. As elaborated on
by Rubin and Rubin (1995):
Life histories focus more on the experiences of an individual and what
he or she felt as he or she passed through the different stages of life.
Life histories can tell us about lifes passages; they can also provide a
window on social change.. .Life histories involve a combination of
narratives and stories that both interpret the past and make it
acceptable, understandable, and important, (p. 27)
It is important to note that the stories that the youths told in this study involved their
very real perspectives about the happenings in their lives. Establishing whether their
descriptions are completely accurate is not the point. Schaffher (1999)
unequivocally asserts that listening to the voices of the homeless and runaway
adolescents themselves is the key to helping them exit the streets. In her words
(ibid.):
So, what can be done for teenage runaways? We need to listen to them,
and to all troubled youths, very carefully. They tell us the answers all
the time in their lyrics and poetry and life stories, (p. xvi)
On the significance of using a modified grounded theory approach, this thesis
attempts to build theoretically-based models, grounded in the research findings and
including the extended model of developmental assets for homeless youths and the
framework depicting pathways into initial and subsequent episodes of homelessness.
50


Grounded theory also involves a combination of inductive and deductive approaches
and can draw upon themes revealed through content analysis. As noted by Berg
(2004):
The categories researchers use in a content analysis can be determined
inductively, deductively, or by some combination of both (Strauss, 1987).
Abrahamson (1983, p. 286) indicates that an inductive approach begins
with the researchers immersing themselves in the documents.. ..in order
to identify the dimensions or themes that seem meaningful to the
producers of each message. In a deductive approach, researchers use
some categorical scheme suggested by a theoretical perspective, and the
documents provide a means for assessing the hypothesis, (pp. 272-273)
The currently study has clearly accessed both approaches.
Site of the Study
The primary contact site selected for this study was Urban Peak, a 501(c)(3)
not-for-profit agency serving homeless and runaway youths in Denver, Colorado.
This site was considered of particular value for this study in that Urban Peak has
retained a fairly extensive and complete data base from which to draw participants
(including both current and formerly homeless clients), has remained in contact with
many of the former recipients of its services and programs, and has maintained
extensive liaisons with other agencies and service providers in the community
dedicated to serving the needs of homeless youths in the greater metropolitan area of
Denver. In fact, many study participants had been in contact with this agency within
51


the year preceding their interviews. Other advantages of this setting included its
relatively central location (which attests to its accessibility as an appropriate service
site for city-center homeless adolescents), the fact that it serves members of both
genders (including housing a fairly equal number of males and females at any point
in time), as well as its accessibility to fairly large numbers of homeless youth from
the three predominant racial/ethnic groups in its surrounding region: African-
American, Caucasian, and Hispanic. Individual participants currently receiving
services through this agency included those 17-21 year-olds living at the shelter (n =
9) as well as several 19-22 year-olds living in Urban Peak transitional apartments (n
= 5).
Additionally, two other groups of subjects, representing a wide range of sites,
were approached to participate in the project. To provide a greater range of
experiences and sample of homeless episodes within the Denver area, some of the
young adults recruited for this study (n = 6) were literally living on the streets at
the time of their interviews, were inconsistently living in adult homeless shelters,
and/or were continuing to be served by other agencies as they attempted to transition
into more stable living situations. At the opposite end of the continuum of subjects,
the final group included young adults who were currently, albeit sometimes
tenuously, living in independent living quarters (n = 7).
While the Youth Advisory Panel of currently-homeless youths and the focus
group comprised of service providers met consistently at the Urban Peak location,
52


the individual iri-depth interviews of study participants were most often conducted in
a neutral setting, such as a mutually-agreed upon restaurant or coffee house. This
was expected to provide a more relaxed atmosphere for attaining information of an
in-depth and personal nature as well as an incentive for participation in the study.
However, all of those living in Urban Peaks transitional apartments were
interviewed in their private apartments, and a few of the others were interviewed at
the Urban Peak location for their convenience.
Sample
The original primary sample was expected to include twenty formerly-
homeless young adults, aged 18 to 29, some of whom were to be identified and
selected by Urban Peak staff (due to ongoing contact) as either having successfully
transitioned to independent living in adulthood or as being unsuccessful in making
this transition. Additional subjects were to be selected by the researchers in order to
provide a more representative sample of homeless episodes along a continuum that
reflected a number of transitional stages (see Appendix C for Prochaskas Stages of
Change). However, due to the somewhat chance selection of participants in
transitional stages (such as those living in transitional living apartments), there were
no guarantees that all stages would be represented. Successful transition was to
include such criteria as consistent domicile, educational attainment, steady
employment, and other factors as determined by collaboration with university and
53


Urban Peak staff, as well as with service provider and youth focus groups. The
seven participants described above as living in independent apartments (and, in one
case, a house) were representative of this group (n = 7). Those identified as being
unsuccessful in making this transition were to include currently homeless young
adults who were living on the streets, in crash houses, in shelters, or who otherwise
had no permanent residences. Ultimately, approximately half (n = 14) of the study
participants fell into this sample group. Those in transition were to include those
living in transitional, subsidized housing, of those who seemed to migrate back and
forth between Semi-stable (temporarily housed) versus unstable living arrangements.
The five young adults living in Urban Peak apartments, as well as one young woman
living in an apartment subsidized by Volunteers of America, comprised this sample
subset (n = 6).
A quick calculation adding the numbers of participants in each sample subset
above reveals a total of 27 young adults who took part in individual life history
interviews for this research project. Concerning those living in shelters or literally
on the streets, it should be noted that those who remain homeless, or transition in
and out of homelessness, are often difficult to contact due to the instability of their
living situations. This fact became all too apparent, as it was impossible to locate
five of the participants from shelters and two from the streets for even a second
interview. Subtracting these seven subjects from the total did provide the anticipated
54


core of twenty subjects, equally divided by gender, who were seen for at least two
interviews each.
The range of ages for the actual young adults interviewed was expanded by
one year on each side of the continuum, allowing one emancipated 17-year-old
female and one admittedly immature 30-year-old male (who initially stated he was
29 and who also happened to be part of a homeless couple) to participate in the
project. Table 3.1 (p. 57) provides a complete breakdown of young adult
participants by age group. Concerning measures of central tendency, the median age
of interviewees was 20, and the mean age was 21.2 years. The high concentration of
19 to 21 year olds appeared to be an artifact of the large numbers of young adults in
contact with, and being provided services by, Urban Peak at this stage of their lives.
The absence of 18-year-olds from the primary study appeared to be completely
coincidental.
Other sample criteria were to include young adults with a range of diverse
characteristics, such as equal numbers of males and females (to provide reasonable
comparison of experiences along gender lines), representation (at least one male and
one female each) from each of the three primary racial/ethnic groups in the region
(African-American, Caucasian, and Hispanic), and representation from homosexual
as well as heterosexual young adults. All of these conditions were eventually met.
Although 15 males and 12 females comprised the entire sample, the core of young
adults participating in two interviews each (n = 20) included 10 males and 10
55


females. Concerning racial/ethnic makeup of the sample, almost 60% were
Caucasians (n = 16), a figure that is fairly representative of city center homeless
youths in Denver; this sample subset included 10 males and six females. The second
highest racial representation (also true for Denver city center homeless youths as
well) was African-American (n = 6), with equal numbers of males and females. Two
additional racial/ethnic groups, including one male and one female each, were
Hispanic (n = 2) and Native American (n = 2). One Asian female completed the
sample. It should be noted that the relatively small sample of Hispanic participants
(despite a relatively large general concentration of Hispanics in Metro Denver) is
reflective of the larger population of homeless young people in Denvers city center.
Homeless youths from this racial/ethnic group are most often found in gangs or
intermittently living with relatives in areas near their original neighborhoods. Please
refer to Table 3.1 (next page) for demographics of the primary sample by age,
gender, and race/ethnicity.
56


GENDER RACE/ETHNICITY
AGE Male Female Cauc. Afr. Am. Hisp. Nat Am. Asian.
17 1 1
18
19 4 2 4 1 1
20 4 3 3 2 1 1
21 3 3 3 1 1 1
22 1 1 1 1
23 1 1
24 1 1
25
26 1 1
27-28
29 1 1
30 1 1
Table 3.1 Demographics of Primary Sample
By age, gender and race/ethnicity.
Descriptive statistics by sexual orientation revealed by far the majority of the
interviewees to be heterosexual, although two gay males and one possibly bisexual
female were interviewed. While this is not a study specifically about race, ethnicity,
or sexual orientation, those factors were considered to serve as sensitizing
concepts (Patton, 1991, p. 391) in both data collection and analysis. Moreover,
one male subject and one female subject were initially selected as additional
representatives from this population to participate in preliminary interviews, or in a
pilot study, that served to guide research design.
Although not considered as sample criteria during the initial study design, a
few additional significant or unique characteristics of the project sample are
noteworthy. First, four of the interviewees were parts of heterosexual homeless
57


couples, who had lived together (either on the streets and/or in apartments) for a
significant period of time. Second, five of the female participants revealed that they
had either given birth or were currently pregnant, and the male partner of one of
these young women claimed paternity for their child currently living in foster care.
One additional young woman had been pregnant but was unable to carry the child to
term. A young Australian woman, who left college in her countiy to see the USA,
found herself pregnant in the youth shelter with no money to return home. Indicative
of the plight of the mentally or intellectually challenged, one developmentally
disabled young man currently living in transitional housing reported having a
difficult time finding acceptance in the work world related to his disability. One
young man who was seen for only one interview was a Columbine High School
survivor, who physically survived the shooting but was desperately fighting for his
life in the wake of three post-Columbine suicide attempts. Still another young man,
living at the shelter at the time of the interview but no longer there for a second
interview the following week, was an ex-Marine kicked out for partying too much.
Finally, one particularly troubling subset of the study sample (n = 7) either was in
jeopardy of losing ground between interviews (that is, being forced from transitional
or independent housing to being back on the street) or actually were again literally
homeless within a few months after participation in the study.
The secondary sample consisted of two focus groups to advise the
researcher regarding study design (including identification of selection criteria) and
58


ultimate construction of a finalized Interview Guide with tentative types of questions
to be posed to study participants. Staff members at Urban Peak helped to identify a
Youth Advisory Panel, comprised of currently homeless youths who were in touch
with Urban Peak at least monthly and many of whom were staying or working at the
facility. Again, attention was given to selection of a fairly representative sample of
participants across gender, racial/ethnic mix, and sexual orientation. Unlike the
primary sample, however, one female panel member in this group identified herself
as being a lesbian. It should further be noted that although this panel was convened
twice, both during the design phase of the project (N = 4) and later after most of the
interviews had been completed (N = 7), two entirely different groups of youths
participated on the two different dates this focus group was utilized, again indicative
of the transient and unstable nature of their living conditions. The Youth Advisory
Panel was specifically organized to provide insights into the nature of the study
population, help guide topic issues and construction of interview questions, and give
feedback (whether on target or not) regarding the studys findings.
The second focus group was comprised of a knowledgeable group of
providers and researchers (N = 8-10) who were members of the Urban Peak
Research Committee, including Urban Peak directors and staff, university faculty
and graduate students, individual service providers (such as physicians and nurses),
researchers from other service agencies, and others dedicated to providing services
to, or promoting informed research regarding, homeless youths. One additional
59


criterion for membership in this focus group was that participants were asked to
indicate particular interest in the parameters of this study, involving the identification
of factors shaping whether homeless youths will or will not be able to successfully
transition to adulthood, as well as to explore study design issues. Feedback was
elicited several times from this group, from questions concerning research design at
the inception of the project, to interpretation of selected study results after the
interviews were completed.
Access and Sample Selection
Access to many participants in the study was negotiated through Urban Peak
in terms of both the primary method (young adults who had formerly been
homeless) and to those participating in the Youth Advisory Council (currently
homeless youths). Using the sample criteria listed above, staff at Urban Peak helped
to identify and locate young adults within their data files to whom they had
previously provided services, and with whom they had contact within the past two
years, or who were currently living in their shelter or transitional living unit. Due to
the importance of establishing trust and legitimacy for the study, as well as for
securing cooperation for participation, initial contacts with many subjects were made
while accompanied by an Urban Peak staff member known to the participants.
In keeping with my research questions, the selection process was purposive
in trying to match the sample to the characteristics of the larger population at Urban
60


Peak with regard to the major demographic characteristics set forth in the study
design: age, gender, race/ethnicity, and sexual orientation. It was also purposive in
trying to find individuals from a variety of living conditions after adolescence. (See
Bernard, 2002, for further discussion of these methods). Although obvious bias was
introduced during this selection process through absence of random sampling, an
attempt was made to attenuate its effects through convenience sampling of subjects
currently living at the shelter. Other young adults were recruited directly from the
downtown streets of Denver (most often near Skyline Park on the 16th Street
Pedestrian Mall), through word of mouth, and in two cases from student contacts of
the researcher at The Metropolitan State College of Denver and The Community
College of Denver on Denvers downtown Auraria campus.
Participants deemed to have made more successful transitions to
adulthood included those residing in the greater Denver metropolitan area in stable,
independently-supported living situations, accessible to Peak staff or the researcher,
and willing to volunteer to take part in two interviews for purposes of the study.
Another group of young adults, identified by either Urban Peak or the researcher as
not having successfully transitioned to independent living, included those currently
living on the streets, at least temporarily or intermittently residing in homeless
shelters, or who had not established permanent domiciles. These participants were
located through homeless shelters, through food lines/soup kitchens in the Denver
area, or literally on the streets and also demonstrated willingness to participate in two
61


interviews during the study. The question Who are you hanging out with?
facilitated a snowball search for additional homeless young adults for potential
participation. Because of the highly sensitive nature of this research, all subjects
read (or had read to them) and signed an appropriate informed consent and
voluntary disclosure form designed to protect the participants, Urban Peak staff,
and the researcher (refer to Appendix D for copies of forms). Demographic
information was voluntarily provided at the start of each interview, with the
informed understanding that individual identities would remain protected and
confidential. Interviewees received $20.00 for each interview, in deference to the
value of their contributions to this research, and were typically provided with
refreshments as well.
Members of the Youth Advisory Panel were identified as currently homeless
youths between the ages of 16 and 21 who had been in contact with Urban Peak
during at least the past month. The majority of participants in this focus group were
currently staying or working at the Urban Peak youth shelter although at least one
was not. Current demographic information on all panel members was collected from
these youths in conjunction with Urban Peak staff. For those under 18 years of age,
an informed consent and voluntary disclosure form was signed by the
participant, with the understanding that co-signatures by parents or legal guardians
might be required unless they were legally emancipated, in keeping with the Human
Subjects Review Board of the University of Colorado at Denver. Since all such
62


minors were noted to either be emancipated or under the guardianship of Urban
Peak, no co-signatures were necessary. All other participants (18 and older)
identified for inclusion in this panel completed and signed their own consent forms.
Finally, members of the knowledgeable provider and researcher focus
group were those known to Urban Peak staff and the researcher, were committed to
working with homeless youths, and had demonstrated willingness to participate in
this study as part of their duties as members of the Urban Peak Research Committee.
Demographic information and signed informed consent forms were collected from
these participants.
While every effort was made to include the most diverse sample of
participants as possible in each group (refer to sections on Sample, In-depth
Interviews, and Youth Advisory Panel), the samples could not be completely
representative of the homeless youth population in Denver, or young adult
population who had formerly been homeless as adolescents, and indeed they were
not intended to be (Patton, 1990, p. 185). Limitations of sample size, site, and point
in time, among other factors, served to preclude this eventuality. In keeping with
both the intent and parameters of qualitative research, I was most interested in
understanding, in depth, the perspectives of a relatively diverse set of formerly
homeless young adults as to what factors or experiences had shaped their ability or
inability to transition into the mainstream of adult society, as well as what they
perceived as the most important support systems and developmental assets that might
63


have allowed this transition to have occurred more easily or even (for those who still
were not living independently) at all.
Pilot Study
As mentioned briefly in the Sample section of this chapter, one male
subject and one female subject were selected from the study population of young
formerly homeless adults to participate in preliminary interviews to help guide
research design. Rubin and Rubin (1995, pp. 203-204) discuss the format and
purpose of these less-structured interviews, including learning enough about the
vocabulary, broader issues, and overall picture of the topical arena to formulate
focused, detailed questions for use in the full-range, in-depth data collection to
follow. This pilot study served to identify some aspects of the research design and
methodology that had been previously missed, to improve and clarify selection of a
number of open-ended questions most pertinent to providing answers to the primary
research questions, and to inform the research regarding vocabulary and issues of
concern to the study population.
In-depth Interviews
The principal venue utilized to collect the widest and most informative range
of data concerning the experiences of young adults who had been homeless as youths
involved 48 in-depth, retrospective interviews of 12 female and 15 male subjects, 17
64


to 30 years of age. In order to better inform differences in successful versus
unsuccessful adaptation to mainstream society into adulthood^ some of these
participants were identified as having successfully completed this process (such as
through completing education or job training, securing steady employment, and/or
residing in a permanent living situation), while others had not yet demonstrated
success in this respect (that is, were either living on the streets or were living in
shelters with absent or unstable employment). Another subset of participants was in
the process of transition, including living in transitional living units or subsidized
housing, with absence of steady and/or self-supporting employment. Following
identification and consent for participation by Urban Peak staff, the latter introduced
many of the prospective participants to me to aid in establishing initial rapport as
well as to set a mutually-acceptable date, time, and place for conducting each
interview. Additionally, I identified a number of other participants to enhance the
range of experiences and progress of subjects within this population.
Interviewing Techniques and Parameters
Open-ended questions, using an interview guide approach (see Appendix
E), were utilized to explore key factors (presence or absence of developmental
assets, family dynamics, personal preferences, physical and mental health issues,
educational and social experiences, employment history, community support
systems, out-of-home placements, nature of homeless experiences) that these
65


participants perceived as either facilitating or hampering their ability to transition
into independent living as young adults. Additional questions attempted to elicit
information about causes of initial or recurring homelessness, adaptation to street
life, concrete details of each subjects current living situation, how these young
adults conceived of time and planning, strengths and weaknesses of interventions,
and advice to service providers, families, and homeless adolescents themselves to
help youths succeed. Often allowing the stories of these young adults to guide the
formulation of other questions pertinent to this study, no two interviews were
ultimately alike. As noted by Patton, The purpose of open-ended interviewing is
not to put things in someones mind.. .but to access the perspective of the person
being interviewed (1990, p. 278). Concerning the use of an interview guide, Patton
(1990) went on to emphasize:
An interview guide is a list of questions or issues that are to be
explored in the course of an interview. An interview guide is
prepared to make sure that basically the same information is obtained
from a number of people by covering the same material. The
interview guide provides topics or subject areas within which the
interviewer is free to explore, probe, and ask questions that will
elucidate and illuminate that particular subject, (p. 283)
While acknowledging that initial design/construction of the interview
instrument was guided by feedback from the pilot interviews, the Youth Advisory
Panel, and the focus group of knowledgeable providers and researchers, some
questions were necessarily modified according to those most appropriate (that is,
66


concerning applicability and vocabulary level) for the specific young adult being
interviewed. A life history was initially collected from each participant in order to
provide context for understanding his or her feelings and perspectives (Rubin and
Rubin, 1995, pp. 27,80). Questions also attempted to elicit differences in
experiences and resilience related to gender, sexual orientation, and culture.
While the original research design called for three interviews with each
young adult, it soon became apparent that it would be difficult to consistently
conduct two, let alone three, interviews with each participant. The rationale for
doing at least two interviews each involved the facts that rapport frequently takes
time to develop, that attention span and fatigue often limit the length of any one
interview, and that sometimes later interviews are richer and more accurate since
interviewees have had time to think about what they are going to say. It was found
that two interviews were actually sufficient in most cases to gamer the necessary
data from each subject. Three interviews were ultimately completed with only one
(the oldest) of the interviewees, who contacted me concerning other thoughts and
recommendations he had been thinking about after his initial two sessions.
Topics Covered
During the first of the two 60-to-90-minute interviews with most participants,
questions focused on provision of demographic information and description of his or
her personal history of homelessness, including factors leading up to the first
67


homeless episode, experiences during periods of homelessness, and periods in which
they did not live on the streets or in shelters. Each of these young adults was
requested to evaluate his or her current living situation and to make connections
between past experiences and this situation. More often during the second interview
he/she was asked to identify their future goals (linked to Prochaskas theories) and to
project what they expected to be doing in the next 30 days, six months, a year, and/or
five years from now. Each subject was also asked to reflect on how support systems
(families, teachers, counselors, service agencies, friends, even street families)
might better aid young people in achieving independent living status as adults. They
were further encouraged to comment on specific programs available through service
agencies that were beneficial, non-beneficial, or needed (that is, should be
implemented) to support youths in this quest.
To foster increased learning between the interviewer and the participants, the
latter were encouraged to return to former topics of interest to them during repeat
interviews. Also, they were specifically queried concerning any changes in status
that had happened to them between these interviews as a measure of the fragility of
domicile or employment in the lives of some of these young people. The importance
of diligent attention to collection of good contact information for each respondent
cannot be over-emphasized in order to set up repeat interviews. This would also
serve to leave the door open to a follow-up study with the young adults in the future.
However, despite this diligence, some of the participants in shelters or literally on the
68


streets had essentially disappeared from their initial points of contact and could not
be found for a second interview.
Acknowledging that success in moving toward independent living in
adulthood may well be a continuum rather than a black and white (successful
versus unsuccessful) distinction, I attempted to ascertain into which stage of
change (see Prochaska, 1990) each participant fit in order to not only further
understand the perspectives of each participant, but also to surmise whether these
young people would remain in their current situations. The possibility that some of
them were at risk for returning to homelessness (in the case of the less
successful group) or that there was cause for optimism for others in moving into
permanent living situations (in the case of the more successful participants) was
considered, and indeed a few from each group did just that within six months
following their interviews.
Taping. Transcription, and Preparation of
Transcripts for Analysis
Concerning how the data were recorded, I audiotape recorded each interview
with the consent of each participant. Immediately following (and occasionally
during) the interview, I further recorded hand-written field notes (Patton, 1990, p.
239) to supplement the tape recordings with context, body language observed, and
other observations pertinent to the discussion, as well as my own reflections
69


regarding the course of the interview. I then submitted the tape recordings for
transcription to a disinterested transcription agency, and carefully proofread and
annotated the transcriptions. To ensure accuracy of the transcriptions, I personally
transcribed some selections and asked for verification of selected segments of some
interviews by a third party (generally someone familiar with the population).
By far the greatest expenditure of time concerning the data collection and
manuscript preparation process involved preparing each interview transcript (often
20 to 30 pages in length) so that it could be imported into the N6 software that was
ultimately used for the qualitative analysis (see Appendix F for sample page). This
included coding some initial demographic data as well as clearly delineating (in
software-required form) the alternating quotations from each speaker (and annotated
explanations) during the interview process. The positive side of this lengthy process
was the thorough immersion of the researcher into every word and context of the
data. This later served to greatly facilitate coding of the significant trends and
findings revealed through the research.
Youth Advisory Panel
As noted earlier in this chapter, Urban Peak staff selected a Youth Advisory
Panel composed of four to seven youths between the ages of 16 and 22. These
youths were homeless at the onset of the study, had been in contact with the Peak
during the past 30 days (including those currently living at the shelter), and were
70


willing to volunteer for the project. In discussing the advantages of forming
collaborations with members of the target population for research purposes, Harper
and Carver relate that although such collaborations require increased time and
commitment, the synergistic knowledge and experience of university researchers,
community-based service providers, and out-of-the-mainstream youth can result in
the development of unique and informative research and programs (1999, p. 250).
Recruited under the criteria listed in the sampling sections, such as diversity
in gender, sexual orientation, and race/ethnicity, this group was accessed twice
during the study, the first time (N = 4) to gather preliminary information to help
guide topical issues, approach, and construction of the interview guide and proposed
procedures prior to the individual interviews. During the second meeting of the
Youth Advisory Panel, following the individual interviews of study participants, an
entirely new set of youths (N = 7) was brought together to elicit feedback regarding
the findings of the study. Table 3.2 (next page) provides demographic information
for both Youth Advisory Panels by age, gender, race/ethnicity, and sexual
orientation.
71


AGE GEN DER RAC E / ETHNICITY SEXUAL ORIENTATION
GrouD One Male Female Cauc. Afr. Am Hisd. Hetero Homo
Under 18 1 1 1
18 1 1 1
19
20 1 1 1
21 1 1 1
Over 21
Gfoud Two
Under 18
18 1 1 1 1 2
19 2 1 1 1 1
20 1 1 1 1 2
21
Over21 1 1 1
Table 3.2 Demographics of Youth Advisory Panels
By age, gender, race/ethnicity, and sexual orientation
Acknowledging that this panel of currently homeless youths served as a focus
group of sorts, Morgans (1997) description of this process seems most succinct:
In combination with other methods, focus groups can provide
preliminary research on specific issues in a larger project or follow
up research to clarify findings from another method, (p. 17)
.. .preliminary focus groups can provide a useful starting point for
individual interviews that involve unfamiliar topics or informants.
The basic idea is to use one or two exploratory focus groups to
reveal the range of the future informants thoughts and experiences
prior to the first individual interview, (p. 22)
Following individual interviews with focus groups allows the
researcher to explore issues that come up only during the analysis
of the interviews. For example, if there appears to be differences
in perspective across different categories of informants, then focus
72


groups can help conform this. (p. 23)
During the first session (90 minutes in length), I introduced myself, briefly
described the purpose of the study, and then proceeded with the introductory inquiry:
This is what I want to know... Besides welcoming general feedback from the
group, I asked such questions as, How should we define success? (versus
failure) and What factors will help determine your success in life? The panel
also reviewed a tentative set of open-ended questions, to be used during the primary
interviews, for relevancy and appropriate wording, and was invited to offer
suggestions for additional questions that might produce the desired types of
responses. Finally, it was deemed important to interview these key informants, who,
as peers, had a unique understanding of the population in the primary study, in order
to identify a number of support mechanisms that these youths felt were, or might be,
most important in helping them move from homelessness into mainstream society.
Questions to be utilized during the follow-up session (also approximately 90
minutes in length, with a short midway break) were based on both the findings of the
prior Youth Advisory session as well as on the results of the in-depth interviews.
Participants were presented with questions to look for congruency with findings of
the interviews with the formerly-homeless young adults, to assess the accuracy of
reported myths concerning homeless youths, and to describe the programs at Urban
Peak that had been helpful or not helpful to them personally or should be considered
73


for adoption. Discussion subsequently ensued with panel members regarding some
of these findings and myths, with education given (where indicated) concerning the
meanings behind the findings as well as the types of programs available through
Urban Peak. Interestingly, there was fairly close agreement between the responses
of these group participants and the young adult interviewees comprising the primary
research sample.
Focus Group of Providers
Again, as mentioned in the Sample and Access and Sample Selection
sections of this chapter, a focus group of knowledgeable providers and researchers
explored study design issues and further informed this investigator regarding the
nature of the study population. As defined by Morgan (1997, p. 2), The hallmark of
focus groups is their explicit use of group interaction to produce data and insights
that would be less accessible without the interaction found in a group. Patton
(1990) further describes the general makeup and advantages of using such a group:
The participants are typically a relatively homogeneous group of
people who are asked to reflect on the questions asked by the
interviewer...The object is to get high quality data in a social context
where people can consider their own views in the context of the views
of others...Focus group interviews also provide some quality controls
on data collection in that participants tend to provide checks and
balances on each other that weed out false or extreme views. The
groups dynamics typically contribute to focusing on the most
important topics and issues, (p. 335)
74


As part of its larger charge as the Urban Peak Research Committee, this
group was initially accessed twice, for approximately 45 minutes during each
session, to discuss design issues as well as possible questions to be used with both
the primary and secondary samples of youths to help guide the interviews. Follow-
up meetings, once formally as a special meeting and five additional times as part of
the Research Committee agenda, were convened regarding logistics, progress, and
validity and elucidation of findings. More generalized notes of pertinent
comments/suggestions from this group were taken by the researcher as ongoing
support for the process.
Data Analysis
The current section will address the analysis of data collected from interview
transcriptions at multiple levels: from the primary sample of young adult participants
and from the two focus groups of currently-homeless youths and knowledgeable
providers/researchers. Briefly, this process progressed from preliminary hand-
coding and annotation (from field notes) of the interview transcripts from all 48 life
history interviews of the primary sample, to review and selection of available
computer software for coding and sorting of open-ended interviews, and finally to
analysis of the data utilizing the software and identifying salient themes .
75


Due to the exploratory nature of this study, I began coding data from
interview transcripts, from focus group transcripts (from the Youth Advisory Panel,
in particular), and from field notes (including observations, perceptions, reflections,
and interpretations of the interviewer) following each individual or group interview,
using existing coding categories suggested by the literature (Strauss & Corbin, 1998;
Rubin & Rubin, 1995; Patton, 1990). As noted by Rubin and Rubin, Coding is the
process of grouping interviewees responses into categories that bring together the
similar ideas, concepts, or themes you have discovered, or steps or stages in a
process (ibid., p. 238). A record of the exact source for each category was kept
(Locke et al., 2000, p. 263) to add credibility to this process. In addition to existing
coding categories, because of the topical areas to be discussed, I often needed to
employ a number of novel coding categories in order to capture the essence of the
trends found in study results dealing with a specific population (i.e., young adults
who had been homeless as youths). The importance of this introductory step was to
inform future analysis through use of a more uniform qualitative data analysis
software package (QSR N6, a 2002 updated version of NUD*IST) in which
researcher-driven coding procedures were imperative.
Following organization of the data into categories, the first step in the
analysis was to analyze the interviews from individual participants, as separate cases,
to begin to develop patterns in specific contexts and to provide a basis for cross-case
analysis. I then began to analyze data across interviews within identified groups (for
76


example, within the group of young adults deemed relatively successful in
transitioning into mainstream adulthood, within the group who had not yet
successfully transitioned into adulthood, and within groups of those judged to be in
transitional stages using Prochaskas model). Attempts were made to identify
themes and patterns among participants within each group as well as any notable
differences among those individuals. The final step involved between-group
analysis, again using pattern coding to identify common themes between groups, but
more importantly attempting to identify significant differences between the groups to
help shed light on what factors or interventions (in keeping with the principal
research questions) influence successful versus unsuccessful transition to
independent adulthood. This was considered particularly relevant for delineation of
factors demarcating gender differences, support systems, or differences in length of
time on the street.
I examined a number of software programs for computer-based qualitative
analysis to ascertain the applicability to and the efficiency that might be gained for
manipulating the large volume of data that was generated during the interviews.
Among those programs initially examined were QSR NUD*IST Version N5,
ATLAS.ti, Ethnograph v5.0, and, somewhat later, QSR NVivo, four packages that
had been used and/or lauded by other qualitative researchers known to this author.
Two software packages were found to be particularly well-designed for the purposes
of this project, that is, qualitative analysis of large amounts of ethnographic (life
77


history) interview data. These were the NUD*IST and ATLAS packages that,
interestingly, were also specifically mentioned by Strauss and Corbin (1998, pp. 275-
79). The ultimate selection of the NUD*IST software reflected ease of use and one
other significant factor. The ATLAS.ti software was designed specifically for
analysis of unstructured data, while the NUD*IST N5 (and, ultimately, N6) software
appeared more compatible with the semi-structured data (generated from a specific
set of open-ended interview questions) collected during this studys in-depth
interviews.
It should be mentioned that the N6 software served merely to facilitate the
coding of the data, thoroughly and systematically. Utilizing the themes and key
concepts I identified through the initial inductive coding of the interview transcripts,
the software increased ease of data retrieval, allowing identification of related
variables among participants. Delineation of the most frequently-occurring and
significant themes emerging from the analysis becomes the subject of the next four
chapters of this thesis.
Also considered of importance for the analysis was to distinguish between
those coding categories of responses generated by the participants themselves and
those that evolved in response to probes or cues given by the examiner. For
example, if a currently homeless young adult participant failed to mention lack of
education or job skills as a possible determining factor in his or her remaining
homeless, I sought to inquire about the possibility of that relationship. Field notes
78


subsequently allowed annotation that a probe was required to elicit the ensuing
response.
Additionally, I looked at the sequence in which issues/factors were brought
up by participants during the interview as well as how often they were brought up
and how they were emphasized or downplayed by informants. Observations
concerning how this ordering of explanations changed over the course of the two
interviews helped to lend further insights into the consistent versus evolving
importance these young adults placed on factors influencing their lives. Further,
increased frequency of similar responses across participants was of particular
importance in coding trends within the data.
One important precaution that was heeded during the analysis was the
importance of not ignoring or discarding any of the data, including that which
run[s] contrary to initial expectations, or that appear[s] not to fit comfortably into
emerging categories for analysis (Locke et al., 2000, p. 263). I attempted to handle
this by illustrating contrary data and atypical responses, as well as by attempting to
extricate individual participant differences that might help to explain these
phenomena (for example, the fact that only two of the participants, who had been
children of Deadheads leading a nomadic existence their entire lives, seemed
comfortable with the idea of living on the streets for periods of time throughout their
lives). Data analysis has also been accompanied by brief but characteristic-revealing
descriptions of each participant (primarily demographic in nature), as well as
79


pertinent questions from each interview, in order to identify and illustrate both
common themes and important distinctions between subjects or groups of subjects.
A final consideration for analysis of tbose transcripts generated from the
group of young adult participants has been to address the presence or absence in their
lives of needs even more fundamental than those listed as outside resources
(external assets) in the Search Institutes Forty Developmental Assets. Factors
such as adequate food to eat, appropriate clothing for climatic changes, affordable
and safe housing, and health care services that youths can afford, easily access, and
feel comfortable using are among the most basic of human needs but are strikingly
missing from the assets list. Possibly at a slightly higher level, but still fairly basic,
are access to money (either earned or received through family and friends), treatment
facilities for mental health or drug and alcohol abuse, and counseling services that
are user friendly and within the means of homeless adolescents and their families.
One goal of the current research has been to help to refine the existing assets model
to include such rudimentary needs, and not assume that they are being met, in order
to make them applicable to all youths (not just those in the middle and upper classes
with stable homes).
Analysis of transcriptions from the focus groups, particularly from the Youth
Advisory Council but also from the focus group of knowledgeable providers and
researchers, has been used primarily as a check to verify the salience of themes and
counter-themes within and across the groups within the study. Important points of
80


consensus or contention during these group discussions were referenced, and
comparisons have been made among the perceptions of the currently homeless
youth, the service providers, and the formerly homeless young adults.
Trustworthiness and Credibility
One of the most frequent objections directed toward qualitative research by
proponents of quantitative research is the relative absence of measures of validity
and reliability. Indeed, these constructs have been the hallmarks of good and
convincing quantitative research. Many researchers have attempted to address the
methodological differences in these terms by utilizing alternative descriptors. As
noted by Rubin and Rubin (1995):
Most indicators of validity and reliability do not fit qualitative
research. Trying to apply these indicators to qualitative work
distracts more than it clarifies. Instead, researchers judge the
credibility of qualitative work by its transparency, consistency-
coherence, and communicability, they design the interviewing to
achieve these standards, (p. 85).
These authors go on to define these constructs. Transparency has to do with the fact
that the reader of a qualitative research report is able to see the basic processes of
data collection. This is achieved through careful transcriptions and use of quotations
from the interviews to support major conclusions. Consistency involves
investigating any inconsistencies (in themes or among individuals) in the research
81


findings, not to eliminate them but to understand why they occurred. Finally,
communicability allows readers to understand and accept the stories that are told and
conclusions that are drawn (even if they do not agree with every interpretation), and
is achieved through the reporting of firsthand experiences demonstrating richness of
detail and abundance of evidence. In reporting the findings, I have relied heavily on
these constructs to ensure credibility and trustworthiness.
In order to minimize threats to validity and to more readily increase the
trustworthiness of this research project, however, I initially took and continued to
review several steps towards those ends. First, concerning credibility, I continued to
verify through my literature search and subsequent literature review the importance
of and need for the current research in that, while the question has been raised by a
number of authors (Robertson & Toro, 1999; James, 1999; and Whitbeck & Hoyt,
1999) as to whether adolescent homelessness is a strong precursor to adult
homelessness, no studies yet have used the proposed methodology to explore this
issue. Moreover, the utilization of focus groups in addition to the in-depth
interviews provides further credibility for the project. This project uses multiple
sources of data and asks the same questions repeatedly in different contexts as
techniques to reduce threats to validity. According to Morgan, The goal of
combining research methods is to strengthen the total project, regardless of which
method is the primary means of data collection (1997, p. 23).
82


In regards to confirmability, I included in my research design several
measures to both verify accuracy of the transcription of data (through review of
selected transcribed segments of the tape-recorded interviews by a third party or by
this investigator) and to provide verification of outcomes/results/trends provided by
the study (through bringing results back to the Youth Advisoiy Panel and to the
focus group of knowledgeable providers and researchers for their feedback).
In an attempt to minimize personal researcher bias, based on potential
preconceived ideas regarding youth success or failure in transitioning to
adulthood as well as factors shaping these transitions, none of the participants was
known to me prior to the interviews, with one exception. One of my Health Care
Management students at The Metropolitan State College of Denver volunteered her
prior homeless status during a classroom discussion regarding decreased access to
health care for some disadvantaged populations. Even in this situation, however,
none of this participants personal history was known to the researcher prior to the
actual interview. Initial selections of many study participants were made by Urban
Peak staff, whose broader knowledge of this population provided balance for this
project. Moreover, my background in providing speech/language therapy and
rehabilitation consulting services to members or potential members of the target
population (such as to foster and group home children and adolescents, to former
runaway youths, and to formerly homeless adults) served to facilitate my ability to
83


establish rapport with these young adults and to navigate the nuances of life history
interviewing techniques.
While it may not be possible to replicate or even safely generalize the results
of this investigation to other populations of homeless and formerly homeless youths
and young adults in other communities, findings and methods identified during the
course of this research should be transferable to other researchers in other locales to
inform and broaden research on this important topic. In this project I used methods
that other qualitative researchers have used with similar populations in other
locations. Additionally, several findings in terms of factors affecting initial
adolescent homelessness and experiences on the streets have been reported in prior
studies.
Concerning future research, risk factors among homeless adolescents most
likely to predict maintenance of homelessness into adulthood, as identified in this
study, may be used for comparison purposes among similar samples. Further, and
possibly even more importantly, the trends and results generated through this
research will better inform policy makers and service providers to homeless or at
risk adolescents regarding improved best practice interventions for increasing
both adolescent developmental assets and potential for successful transition to
adulthood. The expectation that methods identified through this project might serve
as a basis for areas of investigation in other populations (for example, interview
techniques or relevant factors and themes) would be considered an added bonus.
84