Citation
Resiliency as a predictor of students future academic performance and graduation from high school

Material Information

Title:
Resiliency as a predictor of students future academic performance and graduation from high school
Creator:
De Baca, Christine Elaine ( author )
Language:
English
Physical Description:
1 electronic file (58 pages) : ;

Subjects

Subjects / Keywords:
Resilience (Personality trait) in adolescence ( lcsh )
High school students ( lcsh )
High school students ( fast )
Resilience (Personality trait) in adolescence ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
This study evaluated the predictive power of social and emotional factors on students’ academic performance and graduation vs. dropout status. Using a sample of 5,158 students from a large urban school district, the researcher found that three variables, Importance of Education, Money Difficulties, and Bad Feelings, accounted for 4.2% of the variance in Year 1 Attendance, 12.4% of the variance of residual GPA values and 4.4% of the variance of residual cumulative courses failed. Academic measures combined with the resilience measures correctly predicted 79.1% of students who dropped out and 81.6% of students who graduated. The inclusion of the resiliency variables did not increase the accuracy of prediction of dropouts when academic measures were considered alone. Through an examination of the predicted values’ group membership, the researcher found that the resilience measures correctly predicted 8.1% of students who dropped out that were incorrectly predicted to graduate by the academic measures.
Thesis:
Thesis (M.S.Sc.)-University of Colorado Denver
Bibliography:
Includes bibliographic information
System Details:
System requirements: Adobe Reader

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
953836288 ( OCLC )
ocn953836288
Classification:
LD1193.L65 2016m D33 ( lcc )

Downloads

This item has the following downloads:


Full Text
RESILIENCY AS A PREDICTOR OF STUDENTS FUTURE ACADEMIC
PERFORMANCE
AND GRADUATION FROM HIGH SCHOOL
by
CHRISTINE ELAINE DE BACA
B.S. Sociology, Arizona State University, 2007
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Social Sciences
Social Sciences Program
2016


2016
CHRISTINE ELAINE DE BACA
ALL RIGHTS RESERVED
11


This thesis for the Master of Social Sciences degree by
Christine Elaine De Baca
has been approved for the
Social Sciences Program
by
Alan Davis, Chair
Omar Swartz
V. Scott Solberg
April 25th, 2016
in


De Baca, Christine, Elaine (M.S.S., Masters of Social Sciences)
Resiliency as a Predictor of Students Future Academic Performance and Graduation
from High School
Thesis directed by Professor Alan Davis.
ABSTRACT
This study evaluated the predictive power of social and emotional factors on students
academic performance and graduation vs. dropout status. Using a sample of 5,158
students from a large urban school district, the researcher found that three variables,
Importance of Education, Money Difficulties, and Bad Feelings, accounted for 4.2% of
the variance in Year 1 Attendance, 12.4% of the variance of residual GPA values and
4.4% of the variance of residual cumulative courses failed. Academic measures combined
with the resilience measures correctly predicted 79.1% of students who dropped out and
81.6% of students who graduated. The inclusion of the resiliency variables did not
increase the accuracy of prediction of dropouts when academic measures were considered
alone. Through an examination of the predicted values group membership, the
researcher found that the resilience measures correctly predicted 8.1% of students who
dropped out that were incorrectly predicted to graduate by the academic measures.
The form and content of this abstract are approved. I recommend its publication.
Approved: Alan Davis
IV


I dedicate this work to my husband Jason and son Luca, for keeping me grounded and
inspiring me to try to make the world a better place.
v


ACKNOWLEDGMENTS
I would like to extend my sincerest gratitude to Dr. Alan Davis and Dr. V. Scott
Solberg for being such amazing mentors and for their guidance and support throughout
my research journey. Many thanks to my family and friends for all their encouragement
and support. Special thanks to Steve Weigler, Karen Niemi, Melissa Schlinger, Nadine
Lawson, Dave Benoit, Leise Roberts, Leah Rodgers, and all other past and present staff at
ScholarCentric for their tireless commitment to guide students towards valuing education
and graduating from high school.
vi


TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION........................................................1
Statement of the Problem and Background Literature...................1
Purpose of the Study.................................................3
Research Questions...................................................4
Hypothesis...........................................................4
II. LITERATURE REVIEW..................................................5
Early Warning Systems................................................5
The ABC Early Warning Indicator......................................6
Social and Emotional Skills as an Early Warning Indicator..........6
Improved Academic Performance through Social and Emotional Skills.7
Background on the Social and Emotional Skills in this Study.......9
Social and Emotional Factors as an Improved Early Warning Indicator.13
III. METHODS..........................................................16
Research Design and Setting.........................................16
Participants........................................................17
Measures............................................................17
Dependent Variables Academic Performance.......................17
Independent Variables Factor Analysis of Resilience Measures...18
vii


Independent Variables New Resilience Measures Descriptions
19
Correlations..................................................20
Procedures of Data Collection...................................21
Academic Performance Data.....................................21
Resilience Data...............................................22
Missing Data..................................................23
IV. RESULTS.......................................................24
V. DISCUSSION AND CONCLUSION.....................................26
REFERENCES.......................................................30
APPENDIX A: SCHOLARCENTRICS RESILIENCE ASSESSMENT
QUESTIONS.....................................................37
APPENDIX B: FACTOR LOADINGS OF SIX NEW RESILIENCE MEASURES
CONSIDERED FOR THE MULTIPLE REGRESSION IN THIS STUDY
(Based on Principal Components Analysis with Varimax Rotation)....46
APPENDIX C: NEW SUB-SKILLS, QUESTION NUMBERS, AND RELIABILITY
VALUES........................................................48
viii


LIST OF TABLES
TABLE
1. Summary of Intercorrelations, Means, and Standard Deviations for Scores on SEL
Measures and Academic Performance Variables..............................19
2. Data Categorization by Year..............................................21
IX


LIST OF ABBREVIATIONS
ABC.....Attendance, Behavior, and Coursework
CASEL... Collaboration for Academic, Social, and Emotional Learning
DPS.....Denver Public Schools
EWI.....Early Warning Indicator
GPA.....Grade Point Average
SEL.....Social and Emotional Learning
SEM.....Structural Equation Modeling
x


CHAPTERI
INTRODUCTION
Statement of the Problem and Background Literature
One of the most important challenges in the United States today is the need to
contain the nations dropout crisis. According to the U.S. Department of Education,
Institute of Education Sciences, the high school graduation rate is 80% (Stetser &
Stillwell, 2014). This rate is substantially lower for American Indians/Alaska Natives
(67%), African Americans (69%) and Hispanics (73%). Additionally, students from low-
income families have a graduation rate below the national average at 72% (ibid). More
than 5.8 million youth between age 16 and 24 are neither in school nor employed (Burd-
Sharps & Lewis, 2012). With economists predicting that at least 50% of new jobs in the
next ten years will require some postsecondary education, our nation is at serious risk of
falling short of meeting this demand (Georgetown University Center on Education and
the Workforce, 2010).
High school dropouts have a deep impact on our nations economy. Some reports
estimate that a person with a college degree earns an average $1 million more over a
lifetime than a high school dropout (Cheesman Day & Newburger, 2002). Moreover, the
nation suffers a significant loss of tax revenue for every student that drops out. One study
estimates that if all the students in the Class of 2011 had graduated, the nations economy
would have collected an additional $154 billion in tax revenue over these students
lifetimes (Alliance for Excellent Education, 2011). Along these same lines, if the dropout
rate remains the same for the next 10 years, the nations economy will lose $3 trillion in
tax revenues (Balfanz, Fox, Bridgeland, & McNaught, 2009).
1


In addition to economic implications, the dropout crisis has a detrimental impact
on our nations health. Much research illustrates a strong correlation between education
and health, even when factors like income are taken into account. People with more
education have fewer health issues and live longer healthier lives. Studies have shown
that lower levels of education and adverse effects on brain, cognitive, and behavioral
development early in life are closely linked with imperative health outcomes later in life,
including cardiovascular disease and stroke, hypertension, diabetes, obesity, smoking,
drug use, and depression. These conditions account for a major segment of preventable
conditions that cause premature mortality in the U.S. (Miller, Simon, & Maleque, 2009).
Education is also directly linked to employed income; individuals who have more
income are able to afford adequate health care, can make healthier choices for nutritious
foods and exercise, and live in safer homes and neighborhoods. Individuals with lower
income have limited financial resources to support a healthy lifestyle and deal with
unexpected illnesses and other health-related challenges. They also experience greater
levels of stress, which can be a trigger for many serious health-related outcomes (Cutler
& Lleras-Muney, 2006). Increasing high school retention is therefore critical to
promoting social and economic mobility and reducing health related disparities.
Some progress to curtail the dropout crisis is being made. Many states are
achieving improvements in graduation rates and state exam scores, and there are fewer
dropout factories (schools that graduate less than 60% of their students) than there were
ten years ago (Balfanz, Bridgeland, Fox, DePaoli, Ingram, & Maushard, 2014). Some
districts/schools have been relatively successful in improving academic outcomes
(including increased graduation rates) especially for low income and minority students,
2


and there is some consensus in the effectiveness of those similar tactics used within these
schools (Cotton, 2003; Lezotte & Snyder, 2011). School districts are beginning to utilize
more effective practices to address the needs of their students; some states/districts are
getting better at allocating funds toward effective programming and interventions; and
improvements in data collection and analysis are enabling schools and districts to create
effective early warning systems.
Many argue that early warning systems are essential to education reform. These
systems can be a powerful tool, for a student can be identified for being at-risk of
dropping out early on. In fact, research shows that students who drop out of high school
usually do not wake up one day and abruptly make the decision to stop going to school;
rather, they gradually become disengaged through a slow process that begins as early as
elementary school (Bruce, Bridgeland, Fox, & Balfanz, 2011). Students who are not
connected to school at an early age are increasingly at-risk for failing academically in
middle and high school (Blum & Libbey, 2004). This is a significant problem because by
the time students enter the ninth grade, 40-60% of all students are chronically disengaged
from school (Klem & Connell, 2004). Early warning systems are important because much
research also shows that even the most at-risk students can graduate if they are identified
and given the proper supports early enough (Durlak, Weissberg, Dymnicki, Taylor, &
Schellinger, 2011). It is therefore essential for educators to utilize effective early warning
systems so they can identify and plan interventions for at-risk students before they get to
the point where they drop out of school.
Purpose of the Study
Previous research has determined that social emotional factors are fairly strong
3


predictors of future academic performance (Davis, Solberg, De Baca, & Hargrove, 2014).
The purpose of this study is to further examine the relationship between social and
emotional learning skills and academic performance measures, including graduation vs.
dropout. This research could provide schools with an improved early warning system
strategy for identifying at-risk youth and understanding the root cause of why students
are or may likely become at-risk for dropping out of high school. This knowledge will
enable educators to better design appropriate personalized intervention strategies for
those students most in need.
Research Questions
The research questions this thesis will strive to answer are: 1) How well do 8th grade
social and emotional scores predict various high school academic outcomes, including
dropout rates? 2) What is the relationship between academic measures and social and
emotional measures and their ability to predict graduation vs. dropout? 3) Are the
resilience measures able to effectively predict dropouts who are not identified by ABC
measures?
Hypothesis
The researcher expects to find that social and emotional skills are predictive of
students' academic performance and graduation vs. dropout. The researcher also expects
to find a correlational relationship between academic measures and the social and
emotional and measures and that the latter can explain an additional amount of variance
of graduation vs. dropout. Finally, the researcher anticipates that the social and emotional
measures are able to identify some at-risk students that are not correctly identified by the
ABC measures.
4


CHAPTER II
LITERATURE REVIEW
Early Warning Systems
According to the U.S. Department of Education, an effective early warning
system includes the following components: 1) is research- and evidence-based; 2) is
comprised of key data points that are readily available to school districts; and 3) is an
effective predictor of students likelihood to dropout vs graduate without additional
supports or intervention (Jobs for the Future, 2014). Thirty-one states currently use
some version of an early warning system, a jump from 18 states in 2011 (Data Quality
Campaign, 2013). Early warning systems emerged in part due to breakthroughs in the
types of data analyses that enable educators to systematically use key data points to
identify students at-risk of dropping out. Much progress has been made in the last decade
in the development of early warning systems that identify students at-risk of dropping
out, and several factors have contributed to the rise of their use in schools/districts. One
factor is a surge of attention on the nations dropout crisis: school s/districts, along with
communities, government entities, and educational organizations have made it a priority
to identify better systems and approaches for reducing the nations dropout rate (Ibid). A
second factor is the enhancement of student information systems and ready access to
school databases and student data. Thirdly, a growing body of research is providing
strong evidence that early identification of a students likelihood to drop out is possible
and that graduation rates increase when educators use this information effectively in
providing school improvements and targeted interventions (Allensworth & Easton, 2007;
Balfanz, Herzog, & Maclver, 2007; Neild & Balfanz, 2006).
5


The ABC Early Warning Indicator
One of the most widely used early warning indicators (EWI), a key component of
early warning systems, was developed by researchers from the Consortium of Chicago
School Research, the Center for Social Organization of Schools at Johns Hopkins
University, and the Philadelphia Education Fund. This indicator is referred to as the ABC
indicator and is comprised of three key data points: Attendance (defined by missing 20
days or 10% of school days), Behavior (defined by two or more serious behavior
infractions), and Course Performance (defined by an inability to read at grade-level by
third grade; failure in English or math in sixth-ninth grade; a GPA of less than 2.0; two or
more failures in ninth grade courses; and failure to earn on-time promotion to the tenth
grade) (Bruce, et al., 2011, pg. 3; Pinkus, 2008).
Researchers followed students beginning in the 6th and 8th grades, respectively,
and tracked their outcomes through high school. The ABC indicator was found to
effectively identify 60% of dropouts who were marked at-risk in all of the ABC
measures. Additionally, only 29% of students who were marked at-risk in one or more of
the ABC measures went on to graduate from high school (Balfanz, Herzog, & Maclver,
2007). These studies demonstrated the importance of the transition to high school and
also confirmed the predictive power of the ABC indicator. These findings have since
been validated numerous times, including similar longitudinal research studies conducted
in AK, CO, FL, IN, TN, TX and VA (Balfanz, Fox, Bridgeland, & McNaught, 2009).
Social and Emotional Skills as an Early Warning Indicator
Social and emotional factors are another key data point that could act as an EWE
Some research has illustrated that social and emotional skills (specifically levels of
resilience) are predictive of students future academic outcomes. The Collaborative for
6


Academic, Social, and Emotional Learning (CASEL), a prominent organization in this
field with a mission to advance the development of social and emotional learning, defines
social and emotional learning as the process through which children and adults acquire
and effectively apply the knowledge, attitudes, and skills necessary to understand and
manage emotions, set and achieve positive goals, feel and show empathy for others,
establish and maintain positive relationships, and make responsible decisions (What is
social and emotional learning?, 2015). Resilience, a key component of social and
emotional learning, is shown to be a critical facet. Referring to the ability to succeed in
school despite adverse conditions such as poverty or abuse, resilience includes
components such as confidence, a sense of well-being, motivation, an ability to set goals,
relationships/connections, and stress management (Close & Solberg, 2008).
Improved Academic Performance through Social and Emotional Skills
Much research shows that resilience can significantly affect school and life
outcomes for youth, including academic success, even for students who are faced with
great adversity (Benson, Leffert, Scales, & Blyth 1998; Scales, Benson, Leffert, & Blyth,
2000). Furthermore, these skills can be learned, measured, and have lasting effects on
academic performance (Bernard, 2004; Close & Solberg, 2008). Much research has
demonstrated a connection between resilience and academic success. A longitudinal
study by Scales, Roehlkepartain, Neal, Kielsmeier, and Benson (2006) found that higher
levels of resilience are strongly correlated with higher grade point averages (GPAs)
among middle and high school students. These findings hold true over time; students
reporting more characteristics of resilience early in the study had higher GPAs three
years later, compared to students with fewer assets at the start (Ibid).
7


Similarly, in a series of studies conducted by the U.S Department of Education,
Waxman and Huang (1997) assessed inner-city students in the south-central United
States. They found that students who ranked in the 90th percentile on the standardized
tests in mathematics were highly resilient, reporting significantly higher levels of task
orientation and satisfaction, social self-concept, achievement motivation, and academic
self-concept than their counterparts who ranked below the 10th percentile.
Reyes and Jason (1993) designed a study to understand successful high school
students in an inner-city school. Two groups of Latino students were identified as being
either at low or high risk for dropping out of school; all students shared a similar
socioeconomic status, parent-student involvement, and parental supervision. They found
that the low risk students reported strong resiliency, an attribute that the high risk
students were significantly lacking.
Hanson and Austin (2003) conducted a longitudinal study of students in
California and found that nearly every measure of resilience was positively related to
concurrent test scores. The highest increases in test scores occurred in schools where the
students reported high levels of resilience. Moreover, resilience development proved to
be equally beneficial for successive test score improvements in both low and high
performing schools.
Toldson (2008) in his Breaking Barriers study, examined the social, emotional,
and cognitive factors contributing to the academic success of African American males
(n=6000). Four overarching components empirically linked to academic performance
were identified: 1) personal and emotional factors, such as emotional well-being and self-
esteem; 2) family factors, including household composition, parents education and
8


relationship with their children; 3) social and emotional factors, including economic
standing and community involvement; and 4) school factors, relating to their perceptions
of school and relationships with teachers.
Background on the Social and Emotional Skills in this Study
In his work to pinpoint the aspects of resiliency most closely linked to academic
performance, Solberg et al. (1998) identified six key skills as the foundations of
educational resiliency: building confidence, making connections, setting goals, managing
stress, increasing well-being, and understanding motivation. Studies conducted by
Solberg in Milwaukee Public Schools between 1998 and 2004 show that when students
learn about and cultivate these six skills, their school performance improves significantly
(Solberg, Carlstrom, Howard, & Jones, 2007; Bandura, 1986; Cohen & Wills, 1985;
Dohrenwend & Dohrenwend, 1974; Eccles & Wigfield, 2002; Egeland, Carlson, &
Sroufe, 1993; Hobfoll, 1989; Ryan & Deci, 2002; Werner & Smith, 1992). The Solberg
instrument is the social and emotional instrument that will be used in this study.
Academic self-efficacy: This construct refers to the degree to which a student
feels capable of successfully performing school-related tasks (Solberg et al., 1998;
Solberg, OBrien, Villarreal, Kennel, & Davis, 1993). Albert Bandura and his colleagues
found that individuals who possess higher academic confidence beliefs are more likely to
persist when challenged with difficult academic material, perform better during tests, and
perceive negative performance evaluations as challenges to overcome rather than threats
to avoid. As a construct, academic self-efficacy has been consistently found to be
associated with a range of academic outcomes (Schunk & Pajares, 2009) and meta-
9


analysis has established an average medium effect size between academic self-efficacy
and school outcomes (Multon, Brown, & Lent, 1991; Schunk & Pajares, 2009).
Motivation: The model of motivation assessed in the Solberg instrument was
drawn from Deci and Ryans self-determination theory. In their model, motivation
includes both extrinsic motivation and self-determined, or intrinsic motivation. Extrinsic
motivation refers to performing an activity because one feels forced to do it or because
one is concerned with disappointing others; extrinsic motivation results in a person
performing the activity in order to avoid sanctions or guilt. Intrinsic motivation, on the
other hand, involves choosing to perform the behavior because it is perceived as
meaningful or enjoyable (Close & Solberg, 2008; Ryan & Deci, 2002, 2008). It has been
well established that youth develop self-determination when educators create a
personalized, caring, and relational classroom environment in which the youth receive
choices and engage in autonomy-supportive interactions (Deci, Schwartz, Sheinman, &
Ryan, 1981).
Connections: This construct is based on a tremendous amount of research that
links the quality of social support systems to development and health. Most notably,
research has indicated that perceived availability of social support consistently provides
health benefits during times of stress. One theory argues that during times of high stress,
social support acts as a buffer to protect one from becoming ill. Another theory argues
that social support enhances ones overall health regardless of stress level (Cohen &
Wills, 1985; Dohrenwend & Dohrenwend, 1974). In a recent study of anxiety among
African American youth, Lewis, Byrd, & Ollendick (2012) found that stress was
10


associated with increased anxiety and availability of social support was associated with
lower anxiety.
Importance of School: This construct measures how well students value education
and whether they feel education will help them achieve their desired life and career goals
(Eccles & Wigfield, 2002). Research has demonstrated that youth attain higher academic
performance when they perceive the subject matter as relevant to and supportive of their
future life goals (Eccles, 2005; Harackiewicz, Durik, Barron, Linnenbrink-Gracia, &
Tauer, 2008; Hulleman, Godes, Hendricks, & Harackiewicz, 2010; Hulleman &
Harackiewicz, 2009). Goal setting has been shown to result in better health and well-
being at later phases of life. One longitudinal study conducted by Paul Baltes and his
colleagues found that healthier life outcomes were related to individuals who engaged in
three goal-setting strategies. These strategies form the title of the researchers SOC
(selection, optimization, compensation) model. Individuals who had higher SOC ratings
selected a few goals, optimized their opportunities to achieve those goals, and
compensated by switching or modifying goals when faced with adversity (Baltes, 1997).
Stress: Noted psychology professor Stevan Hobfoll argued that stress can be
understood as ones ability to conserve emotional, psychological, and behavioral
resources (Hobfoll, 1998). Research has consistently found a very strong correlation
between academic confidence and academic stress. This means that individuals with
stronger academic confidence have the personal resources they need to manage the
pressures associated with performing academic-related tasks (Solberg, et al., 1998; Torres
& Solberg, 2001; Solberg & Villareal, 1997).
11


Literature has highlighted the risk of coexistence among emotional, psychological,
and academic dimensions in which a student with problems in one domain is more
inclined to concurrently have difficulties in other domains, such as academic
performance. In contrast, youth with positive academic outcomes and minimal
psychological distress tend to be better adjusted (Valdez, Lambert, & Ialongo, 2011).
Furthermore, research has demonstrated that students with low stress were more likely to
continue their education and had higher GPAs than those that experienced higher levels
of stress (Dornbusch, Mont-Reynaud, Ritter, Chen, & Steinberg, 1991; Gillock & Reyes,
1999; Windle & Windle, 1996).
Health and Well-being: This construct is closely linked to stress; many
cumulative risk factors affect health and well-being. For youth living in lower-income
communities, cumulative risk factors include access to health care during neonatal
development, birth, and childhood; quality of housing; and level of community violence.
Living in situations characterized by high cumulative risk can result in chronic stress and
health concerns. Some implications of this include increased drug use, risky sexual
activity, and school failure (Evans, 2004; McEwen, 1998). Additionally, a longitudinal
study, in which student mental health was examined by well-being and psychological
distress indicators, found that students with a combination of low well-being and higher
levels of distress were most at risk for low GPAs (Suldo, Thalji, & Ferron, 2011).
In previous research, structural equation modeling was used to evaluate these
social and emotional learning factors in relation to both one another and academic
outcomes. Students who stated they had strong connections to teachers and peers also
stated that school was more meaningful, enjoyable, and relevant to their future goals.
12


Higher academic self-efficacy was associated with students belief that education was
relevant and also with a higher sense of well-being. The combination of academic
motivation, academic self-efficacy, and health management was associated with
recording better academic outcomes during their initial entry into the ninth grade (Close
& Solberg, 2008). An extension of this research illustrated that academic self-efficacy
mediates the relationship between motivation and academic stress and that perceived
importance of education was predicted by a combination of academic self-efficacy and
motivation (Gillis, 2011).
Social and Emotional Factors as an Improved Early Warning Indicator
In summary, while much research shows that attendance, behavior, and
coursework are strong predictors of students future academic success, a growing number
studies show that social and emotional skills are also a key factor in whether students
drop out or graduate (Close and Solberg, 2008; Eccles, 2005; Benson et al., 1998).
Previous research has demonstrated that resilience skills assessed at the end of middle
school can effectively predict future high school academic success and serve as an early
warning indicator for future dropout risk (Davis et al., 2014).
Combining the ABC indicator with a social and emotional index can provide an
improved EWI for a couple of reasons. First, research has shown that the ABC indicator
does not identify all students who are or may become at-risk for dropping out. In the
same study that found the ABC indictors to effectively predict 60% of dropouts,
researchers also found that nearly one-fourth of the students who went on the drop out of
high school were not marked at-risk by any of the ABC measures (Balfanz, Herzog, &
Maclver, 2007). Social and emotional measures may be able to identify some of these
13


students since they identify underlying issues that may not yet be manifesting in the
students academic performance.
Second, social and emotional data is not only predictive of students likelihood to
drop out, but it also provides educators with the root causes of why students are
disengaged and struggling. For example, two students might be failing math for very
different reasons: the first might simply be struggling with the concepts and might need
more one-on-one tutoring; the second might be dealing with money difficulties at home
and skips school to provide income for the family. The interventions for these two
students are very different, yet access to the ABC indicator alone does not shed light on
why the students are failing. Only when the educator understands the students social and
emotional skills can he/she provide the students with the proper supports and tailored
interventions.
A third limitation of the ABC indicator is that it is only retroactive it is solely
based on students past academic performance. If academic data is unavailable (i.e. due to
an out-of-district transfer or inability to obtain a students academic record), then the
district loses the benefit of the predictive value of the ABC indicator. But if the school
administered a social and emotional assessment when the student enrolled in the district,
educators could be made aware of any underlying issues that could contribute to
disengagement, academic failure, and dropping out. Because social and emotional skills
can be measured at any point in time, this indicator therefore has more flexibility in its
ability to act as a readily accessible data point a key element of an effective early
warning system.
More research is needed to examine the relationship between social and emotional
14


skills and academic performance, specifically as they relate to and are able to predict
whether students graduate or drop out of high school. Schools could benefit from an
improved early warning indicator that 1) more accurately identifies students at-risk for
dropping out of high school; and 2) provides an improved ability to identify root causes
of why students are struggling, which would enable educators to design more appropriate
and personalized interventions.
15


CHAPTER III
METHODS
Research Design and Setting
This study will utilize academic data from Denver Public Schools (DPS) for three
cohorts of students who entered the ninth grade in 2008, 2009, and 2010. Students social
and emotional data was obtained through ScholarCentrics validated resilience
assessment, which measures students perception on the importance of education;
confidence levels; sense of well-being; connections with family, teachers, and peers;
levels stress; and whether they are motivated to learn (Close & Solberg, 2008). The
assessment is based on 20 years of university research led by Dr. Scott Solberg (currently
at Boston University) and a team of researchers at the University of Wisconsin. To date,
approximately 130,000 students have taken the resilience assessment across the U.S. DPS
has utilized ScholarCentrics resilience assessment as a district-wide initiative since
2007. The district currently uses the data within both the middle and high schools as part
of their early warning system to help identify at-risk students and target appropriate
wrap-around services and interventions.
The researcher will link the students resilience scores with academic
performance measures captured by DPS. Academic indicators include: grade point
averages (GPAs), attendance, behavior incidents, cumulative courses failed, 9th and 10th
grade state exam scores, and graduation vs. dropout status. The researcher will analyze
the relationships using linear regression and logistic regression.
The researcher has been an employee of ScholarCentric since 2008; the company
granted access to the resiliency data for the purposes of this research. The academic data
16


was obtained through a formal research request through DPS. This thesis research was
approved by the Colorado Multiple Institutional Review Board (COMIRB).
Participants
This study will utilize a sample from three cohorts of DPS students which consists
of a total of 5,158 students who completed ScholarCentrics resiliency assessment as
incoming ninth graders in 2008 (n=1295), 2009 (n=1701), and 2010 (n=2103). The
sample is comprised of the following race/ethnicity groups: 51% Hispanic, 20% White,
19% African American, 6% Native Hawaiian or Other Pacific Islander, 3% Asian and 1%
American Indian or Alaskan Native. There were 48% males, and 64% of the students
qualified for free or reduced lunch.
Measures
Dependent Variables Academic Performance
A number of high school academic performance indicators were employed as
dependent variables and include: exit code, GPAs, attendance, cumulative failed courses,
behavior incidents, and state exam scores. All three cohorts (2008, 2009, and 2010) had
academic data for at least four years of high school.
Exit Code. Using DPSs Exit Code descriptions, the researcher recoded the Exit
Codes into the following categories: 1) transferred out of the district; 2) still in school; 3)
dropped out and did not return; 4) dropped out but returned to graduate; and 5) graduated.
DPS reported 3,680 students who graduated, 436 who dropped out, and 70 who returned
to graduate after having previously dropped out. Due to time limitations, this study will
only focus on students who either graduated or dropped out and did not return.
17


GPAs. Cumulative GPA data was provided by year and was measured on a 0-5 scale
(with 5 given for an A in an Advanced Placement or IB course). Each years GPA
reflects the total GPA for all high school courses completed. For instance, Year 1
includes just 9th grade; Year 2 includes the cumulative GPAs of Years 1 and 2; etc.
Attendance. Cumulative Failed Courses. Behavior Incidents, and State Exam Scores.
Attendance data is listed as percentage of time attended and includes rates for each year.
Cumulative courses failed were reported as one variable to include courses failed in 9th-
12th grades. Behavior incidents include both expulsions and number of in- and out-of
school suspensions. State exam scores include exam scores for Math, Reading, and
Writing for the 9th and 10th grades. These measures were continuous scale scores, which
were vertically equated to allow them to serve as a measure of growth from grade to
grade.
Independent Variables Factor Analysis of Resilience Measures
The ScholarCentric assessment includes a total of 108 questions, which are grouped
into six resilience factors and 18 subscales. The researcher wanted to test whether the
subscales reliability and their predictive power would increase with revised subscale
categorizations. The researcher thus ran exploratory factor analysis on the items in each
of the original six factors. To do this, the six factors were separated into 2 groups:
Group A included the items from: Education, Connections, and Stress
Group B included the items from: Confidence, Well-being, and Motivation
The items were analyzed using principal components analysis with varimax rotation as an
initial step in examining construct validity by demonstrating that the items are highly
correlated with other items in the same subscale, and less correlated with items in other
18


subscales. Reliability was evaluated using Cronbach alpha estimates of internal
consistency. The rotated factor solution identified thirteen factors, with alpha reliabilities
ranging from .679 (one scale) to .93. Six of the new factors were selected for analysis due
to their high alpha reliabilities: Money Difficulties, Ability to Perform in School,
Education, Bad Feelings, Intrinsic Motivation, and Family Connections (see Appendix A
for the full survey questions; Appendix B for the factor loadings of the new skills; and
Appendix C for the reliability values of the new resilience measures).
Independent Variables New Resilience Measures Descriptions
Money Difficulties. This measure contains 5 items that assess the degree to which
students have difficulty paying for items they need (i.e. food, school supplies). Responses
range from 1 {almost never) to 5 {almost always), with higher scores indicating a greater
degree of difficulty. Internal consistency coefficient for the total scale using Cronbachs
alpha was .89.
Ability to Perform in School. This measure contains 9 items that assess the degree to
which students have difficulty performing academic-related tasks (i.e. taking tests,
completing homework on time). Responses range from 1 {almost never) to 5 {almost
always), with higher scores indicating a greater degree of difficulty. Internal consistency
coefficient for the total scale using Cronbachs alpha was .883.
Education. This measure contains 10 items that assess the degree in which students
value education and believe school and college are important to their future (i.e. it is
important that I finish, go to college, etc.). Responses range from 1 {strongly disagree) to
5 {strongly agree), with higher scores indicating a greater value of education. Internal
consistency coefficient for the total scale using Cronbachs alpha was .93.
19


Bad Feelings. This measure contains 14 items that assess how often students have
negative health-related experiences with manifestations in aggression, physical
symptoms, or depressed feelings (i.e. breaking things when angry, headaches, feeling
hopeless). Responses range from 1 {almost never) to 5 {almost always). Note that
ScholarCentric flips the student responses within this construct to make the reports that
the company provides to schools easier to read. This study utilizes the flipped scores,
which mean that higher scores indicate a higher sense of well-being and lower scores
indicate a lower sense of well-being. Internal consistency coefficient for the total scale
using Cronbachs alpha was .921.
Intrinsic Motivation. This measure contains 7 items that assess whether students go to
school because they find it meaningful (i.e. the reason I keep coming to school is because
I really enjoy school, because education is important for the goals I have, etc.). Responses
range from 1 {strongly disagree) to 5 {strongly agree), with higher scores indicating a
higher intrinsic motivation. Internal consistency coefficient for the total scale using
Cronbachs alpha was .893.
Family Connections. This measure contains 5 items that assess the degree to which
students feel supported by their family members (i.e. I am very close to at least one
member of my family). Responses range from 1 {strongly disagree) to 5 {strongly agree),
with higher scores indicating stronger support from family. Internal consistency
coefficient for the total scale using Cronbachs alpha was .827.
Correlations
Correlations were run on the new resilience measures described in section 3.3.b.
and the academic performance measures (3.3.a.). Table 1 reports the means, standard
20


deviations, and correlations among the variables. Cumulative GPA and State Exam
Scores generally had the highest correlations of the academic variables.
TABLE 1
Summary of Intercorrelations, Means, and Standard Deviations for Scores on SEL
Measures and Academic Performance Variables
Measure 1 2 3 4 5 6 7 8 9 10 11 12
1. Money Difficulties -
2. Ability to Perform in School .577 -
3. Education -.186 -.174 -
4. Bad Feelings -.492 -.569 .167 -
5. Intrinsic Motivation -.178 -.257 .482 .197 -
6. Family Connections -.260 -.281 .418 .285 .395
7. Cumulative GPA -.111 -.213 .206 .169 .163 .137 -
(Year 4) 8. 9th Grade State -.211 -.232 .168 .170 .087 .095 .635
Exam-Math 9. 9th Grade State -.237 -.111 .201 .154 .109 .132 .651 .751
Exam-Reading 10. 9th Grade State -.237 -.226 .201 .154 .109 .132 .651 .751 1.000
Exam-Writing 11. Cumulative -.118 -.084 .111 .108 .078 .037* .531 .252 .234 .234
Attendance 12. Cumulative Courses Failed .151 .150 .129 .132 .111 .075 .720 .397 .378 .377 .507 -
M 1.72 2.30 4.52 4.02 3.94 4.35 2.85 556.66 550.09 549.98 .85 4.86
SD .84 .83 .59 .81 .80 .70 .96 67.88 70.49 70.47 .14 6.69
*Correlation is significant at the 0.05 level (2-tailed).
All other values are significant at the 0.01 level (2-tailed).
Procedures of Data Collection
Academic Performance Data
Academic performance data was provided by DPS and includes the following
variables: GPAs, behavior incidents, cumulative courses failed, attendance, state exam
scores, and graduation vs dropout status. This data was provided for up to five years of
21


the students enrollment in high school grades nine, ten, eleven, twelve, and fifth year
twelfth. Due to time limitations, this study will only include data for grades nine through
twelve, decision made based on research showing that the majority of students who
graduate do so on time (Neild & Balfanz, 2006).
The researcher matched the resilience and academic data by district-assigned Student
IDs and then stripped the file of the ids to maximize the protection of student privacy. To
account for students who repeated one or more grades, the researcher categorized the data
by year, rather than by grade. Table 2. illustrates this categorization system. Eighth grade
year indicates the year the student ended the 8th grade (i.e. Cohort 1 was in 8th grade
during the 2007-2008 school year), and Year 1 indicates the year the student first entered
high school. Years 2-5 indicate the subsequent years the student was enrolled in high
school.
TABLE 2
Data Categorization by Year
8th grade Year Year 1 Year 2 Year 3 Year 4 Year 5
Cohort 1 2008 2009 2010 2011 2012 2013
Cohort 2 2009 2010 2011 2012 2013 2014
Cohort 3 2010 2011 2012 2013 2014 2015
Resilience Data
Resilience data was collected by ScholarCentric through the companys online
resilience assessment. The survey was administered to students the summer prior to
entering the ninth grade. Teachers and other educators administered the assessment as
part of normal education practice in a normal education setting. At no time did the
researcher have contact with students. The ScholarCentric resilience survey consists of
22


108 questions, is available in English and Spanish, and takes students approximately 30
minutes to complete (See Appendix A for the full survey questions).
Missing Data.
Missing values were not imputed for dependent variables; cases with missing data for
a dependent variable were dropped from the analysis involving that dependent variable.
Missing items within the social emotional subscales ranged from .4% to 4.7% of cases,
depending upon the subscale. In those cases, if two or more items were answered, the
mean of the answered items was substituted for missing items within the subscale. If
fewer than two of the items for a scale were complete, the case was dropped from the
analysis.
23


CHAPTER IV
RESULTS
The analysis employed linear regression and logistic regression to determine the
percent of variance in GPA, cumulative courses failed, and attendance attributable to the
measured social and emotional variables. Due to time and data availability constraints,
the academic measures were limited to these variables.
Bivariate correlations among the independent variables (see Table 1.) were
examined to identify variables with the highest correlations that would contribute to
multicollinearity if used together in a regression analysis. The three variables Money
Difficulties, Importance of Education, and Bad Feelings emerged as the highest social
and emotional bivariate correlates of Year 1 Attendance (r=-.133, r=.145, and r=.159,
respectively). The remaining independent variables were omitted to avoid
multicollinearity. In a linear-regression analysis, these variables explained 4.2% of the
variance in Year 1 Attendance (R=.205, R2=.042, p < .001). Residual attendance values,
in which the variance explained by the resilience measures was removed, were saved.
Linear regression found that residual attendance explained 31% of the variance in Year 1
GPA (R=.553, R2=.306, p < .000) and 18% of the variance in cumulative courses failed
(R=.419, R2=. 175, p < .000). Multiple regression was then used to account for variance
in GPA associated with resiliency after controlling for attendance. The three resiliency
variables (Money Difficulties, Importance of Education, and Bad Feelings) explained
12.4% of the variance of the residual GPA values (R=.353, R2=.124, p < .000) and 4.4%
of the variance of the residual cumulative courses failed (R=.211, R =.044, p < .000).
24


Logistic regression found that Year 1 Attendance and Year 1 GPA correctly
predicted 33.9% of students who dropped out and 98.2% of students who graduated (cut
value=.5, Cox & Snell R2=.196, Nagelkerke R2=.404). These academic measures
combined with the three resilience measures correctly predicted 33.1% of students who
dropped out and 98.3% of students who graduated (cut value=.5, Cox & Snell R2=.192,
Nagelkerke R2=.404). When the cut value was increased to .9 (to err on the side of over-
identifying students predicted to drop out), logistic regression found that the academic
measures correctly predicted 79.7% of students who dropped out and 80.8% of students
who graduated (Cox & Snell R2=.196, Nagelkerke R2=.404). Using this same cut value,
the academic measures combined with the resilience measures correctly predicted 79.1%
of students who dropped out and 81.6% of students who graduated (Cox & Snell
R2=.192, Nagelkerke R2=.404). The inclusion of the three resiliency variables did not
increase the accuracy of prediction of dropouts or graduates.
The predicted values group memberships were saved in order to determine if the
resilience measures accurately identified dropouts that were not correctly identified by
the academic measures. The resilience measures correctly identified 8.1% of students
who dropped out and had been incorrectly identified as predicted to graduate by the
academic measures. Both the academic and resilience measures incorrectly identified
12% of dropouts.
25


CHAPTER V
DISCUSSION AND CONCLUSION
Early warning systems are a critical component of schools ability to improve
their graduation rates, and they depend on readily accessible useful/actionable data and
effective EWIs that accurately predict students future academic performance. This study
evaluated 1) whether 8th grade resilience measures are predictive of various high school
academic outcomes, including dropout rates; 2) the relationship between academic
measures and social and emotional measures and their ability to predict graduation vs.
dropout; and 3) whether the resilience measures are able to effectively predict dropouts
who are not identified by ABC measures.
The three social and emotional measures Money Difficulties, Importance of
Education, and Bad Feelings were found to be significant predictors of Year 1 GPA and
cumulative courses failed, after controlling for the amount of variance explained by these
resilience measures on attendance. Attendance was selected as the academic benchmark
because of the immense amount of research linking this measure with students academic
success (Bruce, et al., 2011; Pinkus, 2008). Attendance is so important because this
measure is highly correlated with a variety of other academic variables including GPAs,
state exam scores, fail rates, dropping out, etc. (Balfanz, Herzog, & Maclver, 2007). In
short, a missed day of school is a missed opportunity to learn.
Behavior incidents were not included in the regression analysis due to a large
amount of missing data on this variable. Due to time constraints, the researcher focused
on GPA, cumulative failed courses, attendance and graduation vs. dropout status. It
would be interesting to revisit the analysis and include state exam scores, since these had
26


some of the highest correlations of the academic variables with the resilience measures.
GPA and attendance alone correctly predicted 79.7% of students who dropped out and
80.8% of students who graduated; these findings are consistent with earlier research
mentioned in the Literature Review that established the ABC indicators as predictive of
future academic outcomes (Bruce, et al., 2011; Pinkus, 2008). This amount of predictive
power was retained when the three resilience measures were combined with GPA and
attendance: the combined measures correctly predicted 79.1% of students who dropped
out and 81.6% of students who graduated. These results were dependent on a cut-value of
.9; the researcher made the decision to err on the side of over-predicting dropouts because
the risks of false positives outweigh the risk of under-identifying students at-risk of
dropping out. Through an examination of the predicted values group membership, the
researcher discovered that the resilience measures correctly predicted 8.1% of students
who dropped out that were incorrectly predicted to graduate by the academic measures.
This suggests that the resilience measures add additional predictive value in that they are
able to identify some students who are at-risk of dropping out but fell under the radar of
the academic measures. A total of 12% of students who dropped out were not correctly
identified by either the academic or resilience measures. This information reveals that
there are other elements beyond academic and social and emotional factors that are
contributing to student dropouts.
In addition to the ability to effectively predict academic performance and
graduation vs. dropout status, resilience measures provide educators with information on
the root cause behind students disengagement and likelihood of dropping out. This can
be powerful information because it can enable educators to better understand students
27


individualized needs and create customized interventions accordingly. Further, because
resilience is malleable, educators can improve academic outcomes through SEL
instruction. Resilient students are more self-regulated and engaged learners, and they
perceive the value of education on their future goals (Zimmerman, 2011; Hulleman et al.,
2010; Hulleman & Harackiewicz, 2009).
The reader should be careful to not generalize the results of this correlational
study. The sample of this research was limited to a single school district, and replication
with other school districts would strengthen the analysis. In addition, social and
emotional measures were assessed using self-report data, and therefore all limitations
associated with traditional survey research are warranted.
Finally, the analysis may be strengthened by an analysis of structural equation
modeling (SEM) and/or path analysis. A main benefit of SEM (an extension of multiple
regression) is that it is able to examine relationships between latent constructs represented
by multiple measures and isolate observational error from measurement of latent
variables (Lei & Wu, 2007). Path analysis (which falls under the umbrella of SEM) may
offer more explanatory power than the methods used in this project because it is able to
accommodate the existence of multiple dependent variables. Path analysis identifies three
types of effect: 1) direct effects; 2) indirect effects; and 3) total effect. This method may
better explain the complex and intertwined relationships of the factors contributing to
student dropouts because it accounts for the association of one variable with another
mediated through other variables in the model (Streiner, 2005).
In summary, the inclusion of social and emotional measures into schools early
warning systems meet the criteria required for an effective early warning system
28


(research and evidence-based; readily accessible data points; and effective predictor of
graduation vs. dropout status) and may result in a more effective system for a number of
reasons. First, social and emotional factors, combined with GPA and attendance are fairly
strong predictors of students future academic performance and graduation vs. dropout
status. Second, by understanding why students are struggling academically, educators can
more effectively design tailored interventions to address students individual needs. This
study focused on the resilience measures Importance of Education, Bad Feelings, and
Money Difficulties. Although the latter of these factors is largely outside of the schools
sphere of influence, schools can and should provide resources to help students cope with
bad feelings, and schools need to work on providing motivating environments and
persuading students that education has intrinsic value and is not simply a societal
expectation. And thirdly, these factors may be able to accurately identify a percentage of
students at-risk for dropping out of high school who are not marked at-risk by academic
measures.
29


REFERENCES
Allensworth, E., & Easton, J. (2007). What matters for staying on-track and graduating in
Chicago Public High Schools. Chicago: Consortium on Chicago School Research.
Alliance for Excellent Education. (2011). The high cost of high school dropouts: What
the nation pays for inadequate high schools. Washington D.C.: MetLife
Foundation. Retrieved from ERIC database. (ED 537542)
Balfanz, R., Herzog, L., Maclver, D. (2007). Preventing student disengagement and
keeping students on the graduation path in urban middle-grades schools: Early
identification and effective interventions. Educational Psychologist, 42(4), 223-
235.
Balfanz, R. (2009). Putting middle grade students on the graduation path: A policy and
practice brief. Westerville, OH: National Middle School Association.
Balfanz, R., Fox, J. H., Bridgeland, J.M., & McNaught, M. (2009). Grad Nation: A
guidebook to help communities tackle the dropout crisis. Washington, D.C.:
America's Promise Alliance, Civic Enterprises, & Everyone Graduates Center.
Retrieved from ERIC database. (ED 505363)
Balfanz, R., Bridgeland, J., Fox, J.H., DePaoli, J., Ingram, E., & Maushard, M.
(2014). Building a GradNation: Progress and challenge in ending the high school
dropout epidemic 2014 Annual Update. Washington, D.C.: America's Promise
Alliance, Alliance for Excellent Education, Civic Enterprises, & Everyone
Graduates Center. Retrieved from http://gradnation.org/resource/building-
gradnation-progress-and-challenge-ending-high-school-dropout-epidemic-2014.
Baltes, P. B. (1997). On the incomplete architecture of human ontogeny: Selection,
optimization, and compensation as foundation of developmental theory. American
Psychologist, 52, 366-380.
Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory.
Englewood Cliffs, NJ: Prentice Hall.
Benson, P. L., Leffert, N., Scales, P. C., & Blyth, D. A. (1998). Beyond the village
rhetoric: Creating healthy communities for children and adolescents. Applied
Developmental Science, 2(3), 138-159.
Bernard, B. (2004). Resiliency: What we have learned. San Francisco: WestEd.
Blum, R.W., & Libbey, H.P. (2004). School connectedness-Strengthening health and
education outcomes for teenagers. Journal of School Health, 74, 229-299.
30


Bruce, M., Bridgeland, J. M., Fox, J. H., & Balfanz, R. (2011). On track for success: The
use of early warning indicator and intervention systems to build a grad nation.
Baltimore, MD: Johns Hopkins University, School of Education, Everyone
Graduates Center. Retrieved from ERIC database. (ED 526421)
Burd-Sharps, S. & Lewis, K. (2012). One in seven: Ranking youth disconnection in the
25 largest metro areas. Measure of America of the Social Science Research
Council. Retrieved from http://ssrc-static.s3.amazonaws.com/moa/MOA-
One in Seven09-14.pdf.
Cheesman Day, J., & Newburger, E. C. (2002). The big payoff: Educational attainment
and synthetic estimates of work-life earnings. Current Population Reports (pp.
23-210). Washington, DC: U.S. Census Bureau. Retrieved from
https://www.census.gov/prod/2002pubs/p23-210.pdf.
Close, W., & Solberg, S. (2008). Predicting achievement, distress, and retention among
lower-income Latino youth. Journal of Vocational Behavior, 72, 31-42. doi:
10.1016/j.jvb.2007.08.007
Cohen, S., & Wills, T. A. (1985). Stress, social support and the buffering hypothesis.
Psychological Bulletin, 98, 310-357. doi: 10.1037//0033-2909.98.2.310
Cotton, K. (2003). Principals and student achievement: What the research says.
Alexandria, VA: Association for Supervision and Curriculum Development.
Cutler D. & Lleras-Muney, A. (2006). Education and health: Evaluating theories and
evidence. Bethesda, MD: National Bureau of Economic Research, 2006.
Data Quality Campaign. (2013). Supporting early warning systems: Using data to keep
students on track to success. Retrieved from:
http://www.dataaualitvcampaign.org/files/Supporting%20Earlv%20Warning%20
Svstems.pdf.
Davis, A., Solberg, V.S., De Baca, C., & Hargrove, T. (2014). Use of social emotional
learning skills to predict future academic success and progress toward graduation.
Journal of Education for Students Placed at Risk (JESPAR), 79:3-4, 169-182,
DOI 10.1080/10824669.2014.972506.
Deci, E. L., Schwartz, A. J., Sheinman, L., & Ryan, R. M. (1981). An instrument to
assess adults' orientations toward control versus autonomy with children:
Reflections on intrinsic motivation and perceived competence. Journal of
Educational Psychology, 73, 642-650. doi: 10.1037//0022-0663.73.5.642
Dohrenwend, B. S., & Dohrenwend, B. P. (Eds.). (1974). Stressful life events: Their
naturand effects. New York: Wiley.
31


Dornbusch, S. M., Mont-Reynaud, R., Ritter, P. L., Chen, Z., & Steinberg, L. (1991).
Stressful events and their correlates among adolescents of diverse backgrounds. In
Colton, M. E., & Gore, S. (Eds.), Adolescent stress: Causes and consequences
(pp. 111-130) New York,, NY: Aldine de Gruyter.
Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., & Schellinger, K. B.
(2011). The impact of enhancing students social and emotional learning: A meta-
analysis of school-based universal interventions. Child Development, 52(1): 405-
432.
Eccles, J. S. (2005). Subjective task value and the Eccles et al. model of achievement-
related choices. In A. J. Elliot & C. S. Dweck (Eds.), Handbook of competence
and motivation (pp. 105-121). New York, NY: Guilford.
Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual
Review of Psychology, 53(1), 109-132.
doi: 10.1146/annurev.psych. 53.100901.135153
Egeland, B., Carlson, E., & Sroufe, L. A. (1993). Resilience as process. Development
and Psychopathology, 5, 517-528. doi: 10.1017/s0954579400006131
Evans, G.W. (2004). The environment of childhood poverty. American Psychologist, 59
(2), 77-92.
Georgetown University Center on Education and the Workforce (2010). Help wanted:
Projections of jobs and education requirements through 2018. Retrieved from
www9.georgetown.edu/grad/gppi/hpi/cew/pdfs/FullReport.pdf.
Gillis, S. A. (2011). A structural equation model of resiliency in adolescence.
(Unpublished doctoral dissertation). University of Wisconsin-Milwaukee,
Milwaukee, Wisconsin.
Gillock, K. L., & Reyes, O. (1999). Stress, supports, and academic performance of urban,
low-income, Mexican-American adolescents. Journal of Youth and Adolescence,
28(2), 259-282. doi: 10.1023/a: 1021657516275
Grossman M., & Kaestner R. (1997). Effects of education on health. In: The Social
Benefits of Education, J.R. Berman & N. Stacy (Eds). Ann Arbor, MI: University
of Michigan Press.
Hanson, T.L., & Austin, G. (2003). Student health risks, resilience, and academic
performance in California: Year 2 report, longitudinal analyses. Los Alamitos,
CA: WestEd.
32


Harackiewicz, J. M., Durik, A. M., Barron, K. E., Linnenbrink-Garcia, L., & Tauer, J.
(2008). The role of achievement goals in the development of interest:
Reciprocal relations between achievement goals, interest, and performance.
Journal of Educational Psychology, 100, 105-122. doi: 10.1037/0022-
0663.100.1.105
Hobfoll, S. (1989). Conservation of resources: A new attempt at conceptualizing stress.
American Psychologist, 44(3), 513-524. doi:10.1037/0003-066X.44.3.513.
Hobfoll, S. E. (1998). Stress, culture, and community, the psychology and philosophy of
stress. Dordrecht, the Netherlands: Kluwer Academic/Plenum Publishers.
Hulleman, C. S., Godes, O., Hendricks, B. L., & Harackiewicz, J. M. (2010). Enhancing
interest and performance with a utility value intervention. Journal of Educational
Psychology, 102(4), 880-895. doi: 10.1037/a0019506
Hulleman, C. S., & Harackiewicz, J. M. (2009). Promoting interest and performance in
high school science classes. Science, 326, 1410-1412. doi:
10.1126/science. 1177067
Jobs for the Future. (2014). Early warning indicators and segmentation analysis: A
technical guide on data studies that inform dropout prevention and recovery.
Boston, MA.: Jobs for the Future.
Klem, A.M., & Connell, J.P. (2004). Relationships matter: Linking teacher support to
student engagement and achievement. Journal of School Health, 74, 262-273; doi
: 10.1111/j. 1746-1561.2004.tb08283.x.
Lei, P. & Wu, Q. (2007). Introduction to structural equation modeling: Issues and
practical considerations. Educational Measurement: Issues and Practices, 26(3),
33-43. Retrieved from http://ncme.org/linkservid/47EFEB5A-1320-5CAE-
6EC90BFDF 09 AA3 9E/showMeta/0
Lewis, K. M., Byrd, D. A., & Ollendick, T. H. (2012). Anxiety symptoms in African-
American and Caucasian youth: Relations to negative life events, social
support, and coping. Journal of Anxiety Disorders, 26(1), 32-39.
doi: 10.1016/j.janxdis.2011.08.007
Lezotte, L. W., & McKee-Snyder, K. (2011). What effective schools do: Re-envisioning
the correlates. Bloomington, IN: Solution Tree Press.
Low, M.D., Low, B.J., Baumler, E.R., & Huynh, P.T. (2005). Can education policy be
health policy? Implications of research on the social determinants of health.
Journal of Health Politics, Policy, andLaw, 30(6): 1131-62, 2005.
33


McEwen, B. S. (1998). Protective and damaging effects of stress mediators. New
England Journal ofMedicine, 322, 171-179.
Miller, W., Simon, P., & Maleque, S. (2009). Beyond health care: New directions to a
healthier America. Washington, D.C.: The Robert Wood Johnson Foundation
Commission to Build a Healthier America.
Mirowsky, J., & Ross, C.E. (2003). Education, social status, and health. Hawthorne, NY:
Aldine de Gruyter.
Multon, K. D., Brown, S. D., & Lent, R. W. (1991). Relation of self-efficacy beliefs to
academic outcomes: A meta-analytic investigation. Journal of Counseling
Psychology, 38, 30-38. doi: 10.1037//0022-0167.38.1.30
Neild, R. C., & Balfanz, R. (2006). Unfulfilled promise: The dimensions and
characteristics of Philadelphias dropout crisis, 2000-2005. Baltimore, MD:
Center for Social Organization of Schools, Johns Hopkins University.
Neild, R. C., and Balfanz, R., & Herzog, L. (2007). An early warning system.
Educational Leadership 65.2 (2007): 28-33.
Pinkus, L. (2008). Using early-warning data to improve graduation rates: Closing cracks
in the education system. Washington, DC: Alliance for Excellent Education.
Reyes, O., & Jason, L.A. (1993). Pilot study examining factors associated with academic
success for Hispanic high school students. Journal of Youth and Adolescence, 22,
57-71.
Ross C.E., & Mirowsky J. (1999). Refining the association between education and health:
The effects of quantity, credential, and selectivity. Demography, 36(4): 445-60,
1999.
Ryan, R. M., & Deci, E. L. (2002). An overview of self-determination theory. In E. L.
Deci & R. M. Ryan (Eds.), Handbook of self-determination research (pp. 3-33).
Rochester, NY: University of Rochester Press.
Ryan, R. M., & Deci, E. L. (2008). A self-determination approach to psychotherapy: The
motivational basis for effective change. Canadian Psychology, 49, 186-193. doi:
10.1037/a0012753
Scales, P. C., Benson, P. L., Leffert, N., & Blyth, D. A. (2000). Contribution of
developmental assets to the prediction of thriving among adolescents. Applied
Developmental Science, 4(1), 27-46.
34


Scales, P. C., Roehlkepartain, E.C., Neal, M., Kielsmeier, J.C., & Benson, P.L. (2006).
The role of developmental assets in predicting academic achievement: A
longitudinal study. Journal of Adolescence, 29(5), 692-708.Washington, DC:
Robert Wood Johnson Foundation Commission to Build a Healthier America,
2009.
Schunk, D. H., & Pajares, F. (2009). Self-efficacy theory. In K. R. Wentzel & A.
Wigfield (Eds.), Handbook of motivation at school (pp. 35-53). New York, NY:
Routledge.
Solberg, V.S.H., Gusavac, N., Hamann, T., Felch, J., Johnson, J., Lamborn, S., & Torres,
J. (1998). The Adaptive Success Identity Plan (ASIP): A career intervention for
college students. Career Developments Quarterly, 47, 48-95.
Solberg, V. S., OBrien, K., Villareal, P., Kennel, R., & Davis, B. (1993). Self-efficacy
and Hispanic college students: Validation of the College Self-Efficacy Instrument.
Hispanic Journal of Behavioral Sciences, 15, 80-95. doi:
10.1177/07399863930151004
Solberg, V. S. H., & Villareal, P. (1997). Self-efficacy, stress, and social support as
predictors of Hispanic college adjustment. Hispanic Journal of Behavioral
Sciences, 19, 182-201.
Solberg, S., Carlstrom, A., Howard, K., & Jones, J. (2007). Classifying at-risk high
school youth: The influence of community violence and protective factors on
academic and health outcomes, Career Development Quarterly, 55, 313-327.
Stetser, M., & Stillwell, R. (2014). Public high school four-year on-time graduation rates
and event dropout rates: School years 2010-11 and 2011-12. First Look (NCES
2014-391).
Streiner, D.L. (2005). Finding our way: An introduction to Path Analysis. Canadian
Journal of Psychiatry, 50, (2), 115-122. Retrieved from
http://www.unt.edu/rss/class/mike/6810/articles/pathintro.pdf.
Suldo, S. M., & Thalji, A., & Ferron, J. (2011). Longitudinal academic outcomes
predicted by early adolescents subjective well-being, psychopathology, and
mental health status yielded from a dual-factor model. Journal of Positive
Psychology, 6(1), 17-30. doi: 10.1080/17439760.2010.536774
Toldson, I. A. (2008). Breaking barriers: Plotting the path to academic success for school-
age African-American males. Washington, D.C.: Congressional Black Caucus
Foundation, Inc.
Torres, J.B., & Solberg, V. S. H. (2001). Role of self-efficacy, stress, social integration,
and family support in Latino college student persistence and health. Journal of
Vocational Behavior, 59, 53-63.
35


U.S. Department of Education, National Center for Education Statistics, Common Core
of Data (CCD). (2014). Public high school four-year on-time graduation rates and
event dropout rates: School years 2010-11 and 2011-12. (NCES 2014-391)
Retrieved from School Year 2011-12, Preliminary Version la. Retrieved from
http://nces.ed.gov/programs/coe/indicator coi.asp.
U.S. Department of Education, National Center for Education Statistics. (1998-2011).
Public elementary/secondary school universe surveys. Retrieved from
https://nces.ed.gov/ccd/pdf/psu98rgen.pdf.
Valdez, C. R., Lambert, S. F., & Ialongo, N. S. (2011). Identifying patterns of early risk
for mental health and academic problems in adolescence: A longitudinal study of
urban youth. Child Psychiatry and Human Development, 42, 521-538. doi:
10.1007/sl0578-011-0230-9 .
Waxman, H.C., & Huang, S.L. (1997). Classroom instruction and learning environment
differences between effective and ineffective urban elementary schools for
African American students. Urban Education, 32(1), 7-44.
Windle, M., & Windle, R. C. (1996). Coping strategies, drinking motives, and stressful
life events among middle adolescents: Associations with emotional and
behavioral problems and with academic functioning. Journal of Abnormal
Psychology, 705(4), 551-560. doi: 10.1037//0021-843x. 105.4.551
Werner, E., & Smith., R. (1992). Overcoming the odds: High-risk children from birth to
adulthood. New York: Cornell University Press.
What is social and emotional learning? (2015). In CASEL. Retrieved May 3, 2015, from
http://www.casel.org/social-and-emotional-learning.
Winkleby, M.A., Fortmann, S.P., & Barrett, D.C. (1990). Social class disparities in risk
factors for disease: Eight-year prevalence patterns by level of education. Prev
Med, 79(1): 1-12, 1990.
Zimmerman, B. J. (2011). Motivational sources and outcomes of self-regulated learning
and performance. In B. J. Zimmerman & D. H. Schunk (Eds.), Handbook of self-
regulation of learning and performance (pp. 49-64). New York, NY: Routledge.
36


APPENDIX A
SCHOLARCENTRICS RESILIENCE ASSESSMENT QUESTIONS
Importance of Education
This section asks about your beliefs about the importance of school and college. Mark the
number on the answer sheet that best represents your present attitude of opinion.
Remember, this is not a test, and there are no right or wrong answers. The range of
answers is:
1 = Strongly disagree
2 = Disagree
3 = Neutral/undecided
4 = Agree
5 = Strongly agree
Using the scale above, please mark the number on the answer sheet that best shows *the
degree to which you agree with each statement below:*
1. Finish school.
2. Do well in school.
3. Go to college.
4. Do well in college.
5. Make sure my teacher knows that I want to do well in school.
6. Find out about colleges.
7. Learn how to be successful in college.
8. Get good grades in school.
9. Learn how to be successful in school.
37


10. Get a college degree.
Confidence
This section asks for information about the degree of confidence you have in completing
a variety of activities with being a student at your school. Mark the number on the answer
sheet that best represents your present attitude or opinion. Remember, this is not a test,
and there are no right or wrong answers. The range of answers is:
1 = Not confident at all
2 = Mostly not confident
3 = Somewhat confident
4 = Mostly confident
5 = Extremely confident
Using the scale above, please mark the number on the answer sheet that best shows *the
degree to which you feel confident in successfully... *
11. Making new friends at school.
12. Talking to teachers about homework.
13. Taking good notes in class.
14. Writing a paper for English class.
15. Joining a sports activity.
16. Understanding what you read in your schoolbooks.
17. Asking a question in class.
18. Joining an afterOschool club.
19. Correctly figuring out math problems.
38


20. Turning in your assignments on time.
21. Going to class every day.
22. Working on a group class project.
23. Getting along with classmates.
24. Doing well on your tests.
25. Using a computer to write a paper.
26. Using the library.
27. Using a computer to write a paper.
28. Participating in class discussions.
29. Keeping up to date on school work.
30. Preparing for a test.
31. Relaxing during a test.
32. Studying with others for a test.
Connections
This section asks your relationships with family, teachers, and friends. Mark the number
on the answer sheet that best represents your present attitude or opinion. Remember, this
is not a test, and there are no right or wrong answers. The range of answers is:
1 = Strongly disagree
2 = Disagree
3 = Neutral/undecided
4 = Agree
5 = Strongly agree
39


Please indicate, by marking the number on the answer sheet that best represents the dgree
to which you agree with the following statements:
33. There is a family member who I can talk to about important decisions in my life.
34. Members of my family recognize my abilities and skills.
35. There is no one in my family who shares my interests and concerns.
36. I am very close with at least one other member of my family.
37. There is no on in my family with whom I feel comfortable talking about my
problems.
38. I can talk about school issues of concerns with a family member.
39. There are family members I can count on in an emergency.
40. Teachers here care about their students.
41. There is a teacher here I can go see to talk about academic problems.
42. Teachers here respect me.
43. Teachers here are interested in my success.
44. There is a teacher here I can talk to about a personal problem.
45.1 have friends here at school.
46. There are friends I can talk to about important decisions.
47. There is a friend I can depend on for help.
48.1 have no friends I can depend on.
40


Stress
This section asks about the stresses in your life. Mark the number on the answer sheet
that best represents your present attitude or opinion. Remember, this is not a test, and
there are no right or wrong answers. The range of answers is:
1 = Almost never
2 = Not very often
3 = Somewhat often
4 = Very often
5 = Almost always
Please indicate the degree to which you have experienced each of the following in the
PAST MONTH:
49. Difficulty trying to fulfill responsibilities at home and at school.
50. Difficulty trying to meet friends.
51. Difficulty taking tests.
52. Difficulty talking with teachers about schoolwork.
53. A fear of failing to meet family expectations.
54. Difficulty asking questions in class.
55. Difficulty living in the local community.
56. Difficulty understanding how to use the school library.
57. Difficulty handling relationships.
58. Difficulty handling your schoolwork load.
59. Difficulty with classmates treating you differently than they treat each other.
60. Difficulty writing papers for class.
41


61. Difficulty learning how to use computers.
62. Difficulty paying for school supplies.
63. Money difficulties due to owing money to others.
64. Difficulty paying for food.
65. Difficulty paying for recreation and entertainment.
66. Difficulty due to your family experiencing money problems.
67. Difficulty getting your homework done on time.
68. Difficulty because of feeling a need to perform well in school.
69. Difficulty from teachers.
70. Difficulty from classmates.
Well-being
This section asks you about how often you have had any of these health-related
experiences during the past week. Mark the number on the answer sheet that best
represents your present attitude or opinion. Remember, this is not a test, and there are no
right or wrong answers. The range of answers is:
1 = Almost never
2 = Not very often
3 = Somewhat often
4 = Very often
5 = Almost always
Please indicate the degree to which you have experiences each of these during the PAST
WEEK:
42


How often have you experienced...
71. Being tried but unable to sleep.
72. Mood swings.
73. Feelings of danger.
74. Feeling depressed.
75. Feelings of self-doubt.
76. Nightmares.
77. Snacking more than usual.
78. Feeling hopelessness.
79. Sleeping less than usual at night.
80. Getting sick a lot.
81. Overeating.
82. Breaking things when angry.
83. Headaches.
84. Increased heartbeat.
85. Fighting with friends.
86. Feeling cranky.
87. Losing your temper.
88. Feeling jumpy.
89. Not sleeping well.
90. An upset stomach.
91. Inability to sleep.
92. Increased appetite.
43


93. Becoming easily upset.
Motivation
This section asks about your reasons for going to school. Different people have different
reasons for going to school; we just want to know how much you agree or disagree with
each reason given below. Mark the number on the answer sheet that best represents your
present attitude or opinion. Remember, this is not a test, and there are no right or wrong
answers. The range of answers is:
1 = Strongly disagree
2 = Disagree
3 = Unsure/undecided
4 = Agree
5 = Strongly agree
The reason I keep coming to school is...
94. Because I really enjoy school.
95. Because if I didnt, Id feel guilty.
96. So I can make lots and lots of money.
97. Because education is important for the goals I have.
98. So important people in my life wont be disappointed in me.
99. Because its fun.
100. Because I have to; its required.
101. Because I dont want to let others down.
102. Because skills like reading, math, and science are important to me.
44


103. Because if I dont, Ill get punished.
104. Because failing to get my diploma would bother and disappoint me.
105. Because there are a lot of interesting things to do.
106. Because I see the importance of learning.
107. Because, to me, education is important.
108.1 wouldnt be here if I really had a choice about it.
45


APPENDIX B
FACTOR LOADINGS OF SIX NEW RESILIENCE MEASURES CONSIDERED
FOR THE MULTIPLE REGRESSION IN THIS STUDY
(Based on Principal Components Analysis with Varimax Rotation)
Component Matrix
Component
1 2 3 4 5 6
1. Finish school. .751
2. Do well in school. .788
3. Go to college. .811
4. Do well in college. .828
5. Make sure my teacher knows that 1 want to do well in school. .664
6. Find out about colleges. .723
7. Learn how to be successful in college. .827
8. Get good grades in school. .770
9. Learn how to be successful in school. .757
10. Get a college degree. 33. There is a family member who 1 can talk to about important decisions in my life. .806 .781
34. Members of my family recognize my abilities and skills. .693
36. 1 am very close with at least one other member of my family. .722
38. 1 can talk about school issues of concerns with .764
a family member. 39. There are family members 1 can count on in an emergency. 49. Difficulty trying to fulfill responsibilities at home and at school. .690 .709
51. Difficulty taking tests. .718
52. Difficulty talking with teachers about schoolwork. .706
53. A fear of failing to meet family expectations. .636
54. Difficulty asking questions in class. .673
58. Difficulty handling your schoolwork load. .792
60. Difficulty writing papers for class. .714
67. Difficulty getting your homework done on time. .756
68. Difficulty because of feeling a need to perform well in school. .739
46


Component Matrix (Continued)
Component
1 2 3 4 5 6
62. Difficulty paying for school supplies. -.562 .603
63. Money difficulties due to owing money to others. -.565 .538
64. Difficulty paying for food. -.573 .625
65. Difficulty paying for recreation and entertainment. -.594 .616
66. Difficulty due to your family experiencing -.598 .558
money problems. 72. Mood swings. .668
73. Feelings of danger. .658
74. Feeling depressed. .729
75. Feelings of self-doubt. .710
78. Feeling hopelessness. .726
82. Breaking things when angry. .643
83. Fleadaches. .558
84. Increased heartbeat. .645
85. Fighting with friends. .636
86. Feeling "cranky." .703
87. Losing your temper. .694
88. Feeling "jumpy." .577
90. An upset stomach .643
93. Becoming easily upset. 94. Because 1 really enjoy school. .707 .669
97. Because education is important for the goals 1 have. .698
99. Because it's fun. .647
102. Because skills like reading, math, and science .706
are important to me. 105. Because there are a lot of interesting things to do. .727
106. Because 1 see the importance of learning. .789
107. Because, to me, education is important. .780
Factor loadings < .35 are suppressed.
47


APPENDIX C
NEW SUB-SKILLS, QUESTION NUMBERS, AND RELIABILITY VALUES
New Sub-Scale Items Reliability
Money Difficulties questions 62, 63, 64, 65, and 66 0.890
Ability to Perform in School questions 49, 51, 52, 53, 54, 58, 60, 67, and 68 0.883
Education questions 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 0.930
Teacher Connections questions 40, 41, 42, 43, and 44 0.821
Family Connections questions 33, 34, 36, 38, and 39 0.827
Family Connections group 2 questions 35 and 37 0.679
Peer Connections questions 45, 46, and 47 0.814
Bad Feelings questions 72, 73, 74, 75, 78, 82, 83, 84, 85, 86, 87, 88, 90, and 93 0.921
Academic Confidence questions 13,16,19, 20, 24, 29, and 30 0.864
Intrinsic Motivation questions 94, 97, 99,102,105,106, and 107 0.893
Extrinsic Motivation questions 98,100,101, and 103 0.710
Sleeping Problems questions 71, 79, 89, and 91 0.890
Eating Problems questions 77, 81, and 92 0.806
48


Full Text

PAGE 1

RESILIENCY AS A PREDICTOR OF STUDENTS FUTURE ACADEMIC PERFORMANCE AND GRADUATION FROM HIGH SCHOOL by CHRISTINE ELAINE DE BACA B.S Sociology Arizona State University 2007 A thesis submitted to the Faculty of the Graduate School of the Univer sity of Colorado in partial fulfillment of the requirements for the degree of Master of Social Sciences Social Sciences Program 2016

PAGE 2

ii 2016 CHRISTINE ELAINE DE BACA ALL RIGHTS RESERVED

PAGE 3

iii This thesis for the Master of Social Sciences degr ee by Christine Elaine De Baca has been approved for the Social Sciences Program by Alan Davis Chair Omar Swartz V. Scott Solberg April 2 5 th 2016

PAGE 4

iv De Baca, Christine, Elaine ( M.S .S Masters of Social Sciences ) Resiliency as a Predictor o from High School Thesis directed by Professor Alan Davis ABSTRACT This study evaluated the academic performance and graduation vs. dro pout status. Using a sample of 5,158 students from a large urban school district, the researcher found that three variables, Importance of Education, Money Difficulties, and Bad Feelings, accounted for 4.2% of the variance in Year 1 Attendance 12.4% of th e variance of residual GPA values and 4.4% of the variance of residual cumulative courses failed A cademic measures combined with the resilience measures correctly predicted 79.1% of students who dropped out and 81.6% of students who graduated. The inclusi on of the resiliency variables did not increase the accuracy of pre diction of dropouts when academic measures were considered alone. T the researcher found that the resilience measures correct ly predicted 8.1% of students who dropped out that were incorrectly predicted to graduate by the academic measures. The form and content of this abstract are approved. I recommend its publication. Approved: Alan Davis

PAGE 5

v I dedicate this work to my husband Jason and son Luca, for keeping me grounded and inspiring me to try to make the world a better place

PAGE 6

vi ACKNOWLEDGMENTS I would like to extend my sincerest gratitude to Dr. Alan Davis and Dr. V. Scott Solberg for being such amazing mentors and for their g uidance and support throughout my research journey. Many thanks to my family and friends for all their encouragement and support. Special thanks to Steve Weigler, Karen Niemi, Melissa Schlinger, Nadine Lawson, Dave Benoit, Leise Roberts, Leah Rodgers, and all other past and present staff at ScholarCentric for their tireless commitment to guide students towards valuing education and graduating from high school.

PAGE 7

vii TABLE OF CONTENTS CHAPTER I INTRODUCTION ................................ ................................ ................................ ..... 1 Statement of the Problem and Background Literature ................................ .............. 1 Purpose of the Study ................................ ................................ ................................ 3 Research Questions ................................ ................................ ................................ ... 4 Hypothesis ................................ ................................ ................................ ................. 4 II LITERATURE REVIEW ................................ ................................ ......................... 5 Early Warning Systems ................................ ................................ ............................. 5 The ABC Early Warning Indicator ................................ ................................ ........... 6 Social and Emotional Skills as an Early Warning Indicator ................................ ..... 6 Improved Academic Performance through Social and Emotional Skills ............. 7 Background on the Social and Emotio nal Skills in this Study ............................. 9 Social and Emotional Factors as an Improved Early Warning Indicator ................ 13 III METHODS ................................ ................................ ................................ ........... 16 Research Design and Setting ................................ ................................ .................. 16 Participants ................................ ................................ ................................ .............. 17 Measures ................................ ................................ ................................ ................. 17 Dependent Variables Academic Performance ................................ ................. 17 Independent Variables Factor Analysis of Resilience Measures ..................... 18

PAGE 8

viii Independent Variables New Resilience Measures Descriptions ...................... 19 Correlations ................................ ................................ ................................ ......... 20 Procedures of Data Collection ................................ ................................ ................ 21 Academic Performance Data ................................ ................................ .............. 21 Resilience Data ................................ ................................ ................................ ... 2 2 Missing Data. ................................ ................................ ................................ ...... 23 I V RESULTS ................................ ................................ ................................ ............. 24 V DISCUSSION AND CONCLUSION ................................ ................................ .... 26 REFERENCES ................................ ................................ ................................ ........... 30 APPENDIX A : QUESTIONS ................................ ................................ ................................ .............. 37 APPENDIX B : FACTOR LOADINGS OF SIX NEW RESILIENCE MEASURES CONSIDERED FOR THE MULTIPLE REGRESSION IN THIS STUDY (Based on Principal Components Analysis with Varimax Rotation) .......................... 46 APPENDIX C : NEW SUB SKILLS, QUESTION NUMBERS, AND RELIABILITY VALUES ................................ ................................ ................................ ..................... 48

PAGE 9

ix LIST OF TABLES T ABLE 1. Summary of Intercorrelations, Means, and Standard Deviations for Scores on SEL Mea sures and Academic Performance Variables .... ......................... ....................19 2 Data Categorization by Year

PAGE 10

x LIST OF ABBREVIATIONS Social, and Emotional Learning Denver Public Schools

PAGE 11

1 CHAPTER I INTRODUCTION Statement of the Problem and Backgr ound Literature One of the most important challenges in the United States today is the need to Institute of Education Sciences, the high school graduation rate is 80% (Stet ser & Stillwell, 2014). This rate is substantially lower for American Indians/Alaska Natives (67%), African Americans (69%) and Hispanics (73%). Additionally, students from low income families have a graduation rate below the national average at 72 % (ibid) More than 5.8 million youth between age 16 and 24 are neither in school nor employed (Burd Sharps & Lewis, 2012) With economists predicting that at least 50% of new jobs in the next ten years will require some postsecondary education, our nation is at s erious risk of falling short of meeting this demand ( Georgetown University Center on Education and the Workforce, 2010). High school dropouts have estimate that a person with a college degree earns an av erage $1 million more over a lifetime than a high school dropout ( Cheesman Day & Newburger, 2002 ) Moreover, the nation suffers a significant loss of tax revenue for every student that drops out. One study estimates that if all the students in the Class o f 2011 would have collected an addition al $154 lifetimes ( Alliance for Excellent Education, 2011 ) Along these same lines, if the dropout rate remains the same for the next 10 tax revenues (Balfanz, Fox, Bridgeland & McNaught 2009).

PAGE 12

2 In addition to economic implications, the dropout crisis has a detrimental impact relation between education and health, even when factors like income are taken into account. P eople with more education have fewer health issues a nd live longer healthier lives. Studies have shown that lower levels of education and adverse effects on brain cognitive and behavioral development early in life are closely linked with imperative health outcomes later in life, including cardiovascular disease and stroke, hypertension, diabetes, obesity, smoking, drug use and depression. These conditions accoun t for a major segment of preventable conditions that cause premature mortality in the U.S (Miller, Simon & Maleque, 2009). Education is also directly linked to employed income; individuals who have more income are able to afford adequate health care, ca n make healthier choices for nutritious foods and exercise, and live in safer homes and neighborhoods. Individuals with lower income have limited financial resources to support a healthy lifestyle and deal with unexpected illnesses and other health related challenges They also experience greater levels of stress, which can be a trigger for many serious health related o utcomes (Cutler & Lleras Muney, 2006) Increasing high school retention is therefore critical to promoting social and economic mobility and reducing health related disparities. Some p rogress to curtail the dropout crisis is being made Many states are achieving improvements in graduation rates and state exam scores and ther e are fewer dropout factories (schools that graduate less than 60% o f their students) than there were ten years ago ( Balfanz, B ridgeland, Fox, DePaoli, Ingram & Maushard, 2014) Some districts/ schools have been relatively successful in improving academic outcomes (including increased graduation rates) especially for low i ncome and minority students,

PAGE 13

3 and there is some consensus in the effectiveness of those similar tactics used within these schools (Cotton, 20 03 ; Lezotte & Snyder, 2011). S chool districts are beginning to utilize more effective practices to addr ess the need s of their students; some states/districts are getting better at allocating funds toward effective programming and interventions; and improvements in data collection and analysis are enabling schools and districts to create e ffective early warning systems. Many argue that early warning systems are essential to education reform. These systems can be a powerful tool, for a student can be identified for being at risk of dropping out early on. In fact, research shows that students who drop out of high school u sually do no t wake up one day and abruptly make the decision to stop going to school; rather they gradually become disengaged through a slow process that begins as early as elementary school (Br uce, Bridgeland, Fox & Balfanz, 2011). Students who are not connected to school at an early age are increasingly at risk for failing academically in middle and high school (Blum & Libbey, 2004) This is a significant problem because by the time students enter the ninth grade, 40 60% of all students are chronically disengaged from school (Klem & Connell, 2004). Early warning systems are important because much research also shows that even the most at risk students can graduate if they are identified and given the proper supports early enough (Durl ak, Weissberg, Dymni cki, Taylor & Schellinger, 2011). It is therefore essential for educators to utilize effective early warning systems so they can identify and plan interventions for at risk students before they get to the point where they drop out of school. Purpose of th e Study Previous research has determined that social emotional factors are fairly strong

PAGE 14

4 predictors of future academic performance (Davis, Solberg, De Baca & Hargrove, 2014) The purpose of this study i s to further examine the relationship between social and emotional learning skills and academic performance measures, including graduation vs. dropout. This research could provide schools with an improved early warning system strategy for identif ying at risk youth and understanding the root cause of why stud ents are or may likely become at risk for dropping out of high school This knowledge will enable educators to better design appropriate personalized intervention strategies for those students most in need Research Questions The research questions this t hesis will strive to answer are: 1) How well do 8th grade social and emotional scores predict various high school academic outcomes including dropout rates ? 2) What is the relationship between academic measures and social and emotional measures and their ability to predict graduation vs. dropout? 3) Are the resilience measures able to effectively predict dropouts who are not identified by ABC measures ? Hypothesis The researcher expects to find that social and emotional skills are predictive of students' academic performance and graduation vs dropout The researcher also expects to find a correlational relationship between academic measures and the social and emotional and measures and that the latter can expl ain an additional amount of variance of gradua tion vs dropout Finally, the researcher anticipates that the social and emotional measures are able to identify some at risk students that are not correctly identified by the ABC measure s

PAGE 15

5 CHAPTER II LITERATURE REVIEW Early Warning Systems According to the U.S. Department of Education, an effective early warning system includes the following components: 1) is research and evidence based; 2) is comprised of key data points that are readily available to school districts; and 3) is an effective predictor o supports or intervention ( Jobs for the Future, 2014) Thirty one s tates currently use some version of an early warning system a jump from 18 states in 2011 ( Data Quality Campaign, 2013). Early warning systems emerged in part due to breakthroughs in the types of data analyses that enable educators to systematically use key data points to identify students at risk of dropping out. M uch progress has been made in the last decade i n the develo p ment of early warning systems that identify students at risk of dropping out, and several factors have contributed to the rise of their use in schools/districts. One ith communities, government entities, and educational organizations have made it a priority (I bid) A second factor is the enhancement of student information systems and ready access to school databases and student data Thirdly, a growing body of research is providing and that graduation rates increase when educators use this informatio n effectively in providing school improvements and targeted interventions (Allensworth & Easton, 2007; Balfanz, Herzog, & MacIver 2007; Neild & Balfanz, 2006).

PAGE 16

6 The ABC Early Warning Indicator One of the most widely used early warning i ndicators (EWI), a key component of early warning systems, was developed by researchers from the Consortium of Chicago School Research, the Center for Social Organization of Schools at Johns Hopkins University, and the Philadelphia Education Fund. This indicato r is referred to as the ABC indicator and is comprised of three key data points: Attendance (defined by missing 20 days or 10% of school days), Behavior (defined by two or more serious behavior ade level by third grade; failure in English or math in sixth ninth grade; a GPA of less than 2.0; two or more failures in ninth grade courses; and failure to earn on time promotion to the tenth Bruce, et al., 2011 pg. 3; Pinkus, 2008 ). Research ers followed students beginning in the 6 th and 8 th grades, respectively, and tracked their outcomes through high school. The ABC indicator was found to effectively identify 60% of dropouts who were marked at risk in all of the ABC measures. Additionally, o nly 29% of students who were marked at risk in one or more of the ABC measures went on to graduate from high school (Balfanz, Herzog, & MacIver, 2007). These studies demonstrated the importance of the transition to high school and also confirmed the predic tive power of the ABC indicator. These findings have since been validated numerous times, including similar longitudinal research studies conducted in AK, CO, FL, IN, TN, TX and VA ( Balfanz, Fox, Bridgeland & McNaught, 2009) Social and Emotional Skills as an Early Warning Indicator Social and emotional factors are another key data point that could ac t as an EWI. S ome research has illustrated that social and emotional skills (specifically levels of resilience ) are predict ive of future academic o utcomes. The Collaborative for

PAGE 17

7 Academic, Social, and Emotional Learning (CASEL ), a prominent organization in this field with a mission to advance the development of social and emotional learning, de fines social and emotional learning as h which children and adults acquire and effectively apply the knowledge, attitudes, and skills necessary to understand and manage emotions, set and achieve positive goals, feel and show empathy for others, establish and maintain positive relationships, and What is social and emotional learning?, 2015 ). R esilience, a key component of social and emotional learning, is shown to be a critical facet Referring to the ability to succeed in school despite adverse conditions such as p overty or abuse, resilience includes components such as confidence, a sense of well being, motivation, an ability to set goals, relationships/connections, and stress management ( Close & Solberg, 2008 ). Improved Academic Performance through Social and Emoti onal Skills Much r esearch shows that resilience can significantly affect school and life outcomes for youth, including academic success, even for students who are faced with great adversity (Benson, Leffert, Scales & Blyth 1998 ; Scales, Benson, Leffert & Blyth, 2000 ) Furthermore, these skills can be learned, measured, and have lasting effects on academic performance ( Bernard, 2004; Close & Solberg, 2008 ) Much research has demonstrat ed a connection between resilience and academic success. A longitudinal study by Scales R oehlkepartain, Neal, Kielsmeier and Benson (2006 ) found that higher levels of resilience are strongly correlated with higher grade point averages (GPAs) among middle and high school students. These findings hold true over time ; st udents reporting more characteristics of resilience early in the study had higher GPAs three years later, compared to students with fewer assets at the start ( I bid ).

PAGE 18

8 Similarly, in a series of studies conducted by the U.S Department of Education, Waxman and Huang (1997) assessed inner city students in the south central United States. They found that students who ranked in the 90 th percentile on the standardized tests in mathematics were highly resilient, reporting significantly higher levels of task orientation and satisfaction, social self concept, achievement motivation, and academic self concept than their counterparts who ranked below the 10 th percentile. Reyes and Jason (1993) designed a study to understand successful high school students in an inner city schoo l. Two groups of Latino students were identified as being either at low or high risk for dropping out of school; all students shared a similar socioeconomic status, parent student involvement, and parental supervision. They found that the low risk students reported strong resiliency, an attribute that the high risk stud e nts were significantly lacking Hanson and Austin (2003) conducted a longitudinal study of students in California and found that nearly every measure of resilience was positively related to concurrent test scores. The highest increases in test scores occurred in schools where the students reported high levels of resilience. Moreover, resilience development proved to be equally beneficial for successive test score improvements in both low and high performing schools Toldson (2008) i n his Breaking Barriers study, examined the social, emotional, and cognitive factors contributing to the academic success of African American males (n=6000). Four overarching components empirically linked to academi c performance were identified: 1) personal and emotional factors, such as emotional well being and self

PAGE 19

9 relationship with their children; 3) social and emotional factors, in cluding economic standing and community involvement; and 4) school factors, relating to their perceptions of school and relationships with teachers. Background on the Social and Emotional Skills in this Study In his work to pinpoint the aspects of resilie ncy most closely linked to academic performance, Solberg et al. (1998) identified six key skills as the foundations of educational resiliency: building confidence, making connections, setting goals, managing stress, increasing well being and understanding motivation. Studies conducted by Solberg in Milwaukee Public Schools between 1998 and 2004 show that when students learn about and cultivate these six skills, their school performance improves significant ly (Solberg, Carlstrom, Howard & Jones, 2007 ; Band ura, 1986; Cohen & Wills, 1985; Dohrenwend & Dohrenwend, 1974; Eccles & W igfield, 2002; Egeland, Carlson & Sroufe, 1993; Hobfoll, 1989; Ryan & Deci, 2002; Werner & Smith, 1992 ). The Solberg instrument is the social and emotional instrument tha t will be us ed in this study Academic self efficacy : This construct refers to the degree to which a student feels capable of successfully performing school related tasks (Solberg et al., 1998; rien, Villarreal, Kennel & Davis, 1993). Albert Bandura and h is colleagues found that individuals who possess higher academic confidence beliefs are more likely to persist when challenged with difficult academic material, perform better during tests, and perceive negative performance evaluations as challenges to ove rcome rather than threats to avoid. As a construct, academic self efficacy has been consistently found to be associated with a range of academic outcomes (Schunk & Pajares, 2009) and meta

PAGE 20

10 analysis has established an average medium effect size between acade mic self efficacy and school outcomes (Multon, Brown & Lent, 1991; Schunk & Pajares, 2009). M otivation : Th e model of motivation assessed in the Solberg instrument was determination theory. In their model, motivation includ es both extrinsic motivation and self determined, or intrinsic motivation. Extrinsic motivation refers to performing an activity because one feels forced to do it or because one is concerned with disappointing others; extrinsic motivation results in a pers on performing the activity in order to avoid sanctions or guilt. Intrinsic motivation, on the other hand, involves choosing to perform the behavior because it is perce ived as meaningful or enjoyable (Close & Solberg, 2008; Ryan & Deci, 2002, 2008). It has been well established that youth develop self determination when educators create a personalized, caring, and relational classroom environment in which the youth receive choices and engage in autonomy supportive interac tions (Deci, Schwartz, Sheinman & Ry an, 1981). Connections : This construct is based on a tremendous amount of research that links the quality of social support systems to development and health. Most notably, research has indicated that perceived availability of social support consistently p rovides health benefits during times of stress. One theory argues that during times of high stress, social support acts as a buffer to protect one from becoming ill. Another theory argues alth regardless of stre ss level (Cohen & Wills, 1985; Dohrenwend & Dohrenwend, 1974). In a recent study of anxiety among African American youth, Lewis, Byrd, & Ollendick (2012) found that stress was

PAGE 21

11 associated with increased anxiety and availability of social support was associa ted with lower anxiety. Importance of S chool : This construct measures how well students value education and whether they feel education will help them achieve their desired life and career goals (Eccles & Wigfield, 2002). Research has demonstrated that you th attain higher academic performance when they perceive the subject matter as relevant to and supportive of their future life goals (Eccles, 2005; Harackiewicz, Du rik, Barron, Linnenbrink Gracia & Tauer, 2 008; Hulleman, Godes, Hendricks & Harackiewicz, 2010; Hulleman & Harackiewicz, 2009). Goal setting has been shown to result in better health and well being at later phases of life. One longitudinal study conducted by Paul Baltes and his colleagues found that healthier life outcomes were related to indiv iduals who engaged in three goal (selection, optimization, compensation) model. Individuals who had higher SOC ratings selected a few goals, optimized their opportunities to achiev e those goals, and compensated by switching or modifying goals when faced with adversity (Baltes, 1997) Stress : Noted psychology professor Stevan Hobfoll argued that stress can be vioral resources ( Hobfoll, 1998). Research has consistently found a very strong correlation between academic confidence and academic stress. This means that individuals with stronger academic confidence have the personal resources they need to manage the p ressures associated with performing academic related tasks ( Solberg, et al., 1998; Torres & Solberg, 2001; Solberg & Villareal, 1997).

PAGE 22

12 Literature has highlighted the risk of coexistence among emotional, psychological, and academic dimensions in which a stu dent with problems in one domain is more inclined to concurrently have difficulties in other domains, such as academic performance. In contrast, youth with positive academic outcomes and minimal psychological distress tend to be better adjusted (Valdez, La mbert & Ialongo, 2011). Furthermore, research has demonstrated that students with low stress were more likely to con tinue their education and had higher GPA s than those that experienced higher levels of stress (Dornbu sch, Mont Reynaud, Ritter, Chen & Ste inberg, 1991; Gillock & Reyes, 1999; Windle & Windle, 1996). Health and W ell being : This construct is closely linked to stress ; m any cumulative risk factors affect health and well being. For youth living in lower income communities, cumulative risk factors include access to health care during neonatal development, birth, and childhood; quality of housing; and level of community violence. Living in situations characterized by high cumulative risk can result in chronic stress and health concerns. Some implica tions of this include increased drug use, risky sexual activity, and school failure (Evans, 2004; McEwen, 1998). Additionally, a longitudinal study, in which student mental health was examined by well being and psychological distress indicators, found that students with a combination of low well being and higher levels of distress were most at risk for low GPAs (Suldo, Thalji & Ferron, 2011). In previous research, structural equation modeling was used to evaluate these social and emotional learning factors in relation to both on e another and academic outcomes S tudents who stated they had strong connections to teachers and peers also stated that school was more meaningful, enjoyable and relevant to their future goals

PAGE 23

13 Higher academic self efficacy was asso relevant and also with a higher sense of well being. The combination of academic motivation, academic self efficacy, and health management was associated with recording better academic outcomes during their i nitial entry into the ninth grade ( Close & Solberg, 2008) An exten sion of this research illustrated that academic self efficacy mediates the relationship between motivation and academic stress and that perceived importance of education was predicted by a combination of academic self efficacy and motivation (Gillis, 2011). Social and Emotional Factors as an Improved Early Warning Indicator In summary, w hile much research shows that attendance, behavior, and re academic success, a growing number studies show that social and emotional skills are also a key factor in whether students drop out or graduate ( Close and Solberg, 2008; Eccles, 2005; Benson et al 1998 ). Previous research has demonstrated that resilie nce skills assessed at the end of middle school can effectively predict future high school academic success and serve as an early warning indicator for future dropout risk (Davis et al., 2014 ). Combining the ABC indicator with a social and emotional in dex can provide an improved EWI for a couple of reasons. First, research has shown that the ABC indicator does not identify all students who are or may become at risk for dropping out. In the same study that found the ABC indictors to effectively predict 60% of dropouts, researchers also found that nearly one fo u rth of the students who went on the drop out of high school were not marked at risk by any of the ABC measures ( Balfanz, Herzog, & MacIver 2007 ) Social and emotional measures may be able to identify some of these

PAGE 24

14 students since they identify underlying issues that may not yet be manifesting in the Second, social and drop out, but it also provides educators with the root cause s of why students are disengaged and struggling. For example, two students might be failing math for very different reasons: the first might simply be struggling with the concepts and might need more one on one tutoring ; the second migh t be dealing with money difficulties at home and skips school to provide income for the family. The interventions for these two students are very different, yet access to the ABC indicator alone does not shed light on why the students are failing. O nly w he n the educator understands the and emotional skill s can he/she provide the students with the proper supports and tailored interventions A third limitation of the ABC indicator is that it is only retroactive it is solely based on student t academic performance. If academic data is unavailable (i.e. due to an out of district transfer ), then the district loses the benefit of the predictive value of the ABC indicator But if the school administered a social and emotional assessment when the student enrolled in the district, educators could be made aware of any underlying issues that could contribute to disengagement, academic failure, and dropping out. Because social and emotional skills can be measured at any point in time this indicator therefore has more flexibility in its ability to act as a readily accessible data point a key element of an effective early warning system More research is needed to examine the relationship between social and emotional

PAGE 25

15 skills and academic performance, specifically as they relate to and are able to predict whether students graduate or drop out of high school. Schools could benefit from an improved early warning indicator that 1) more accurately identi fies students at risk for dropping out of high school; and 2) provides an improved ability to identify root causes of why students are struggling, which would enable educators to design more approp riate and personalized intervention s.

PAGE 26

16 CHAPTER III METHOD S Research Design and Setting This study will utilize academic data from D enver Public S chools (DPS) for three social and emotional data was obtained through s validated resilience a ssessment, which measures confidence levels; sense of well being; connections with family, teachers, and peers; levels stress ; and whether they are motivated to learn ( Close & Sol berg, 2008 ). The assessment is based on 20 years of university research led by Dr. Scott Solberg (currently at Boston University) and a team of researchers at the University of Wisconsin. To date, approximately 130,000 students have taken the resilience as sessment across the U.S. DPS as a district wide initiative since 2007. The district currently uses the data within both the middle and high schools as part of their early warning system to help identify a t risk students and target appropriate wrap around services and interventions. The researcher will link the student scores with academic performance measures captured by DPS. Academic indicators include: grade point average s (G PAs ) attendanc e, behavior incidents, cumulative courses failed, 9 th and 10 th grade state exam scores, and graduation vs dropout status. The researcher will analyze the relationships us ing linear regression and logistic regression The researcher has been an employee o f ScholarCentric since 2008; the company granted access to the resiliency data for the purposes of this research. The academic data

PAGE 27

17 was obtained through a formal research request through DPS. This thesis research was approved by the Colorado Multiple Insti tutional Review Board (COMIRB). Participants This study will utilize a sample from three cohorts of DPS students which consists of a total of 5, 158 incoming ninth graders in 200 8 (n=1295) 20 0 9 (n=1701) and 20 10 (n=2103) The sample is comprised of the following race/ethnicity groups: 51% Hispanic, 20% White, 19% African American, 6% Native Hawaiian or Other Pacific Islander 3% Asian and 1% American Indian or Alaskan Native. There were 48% m ales, and 64% of the students qualified for free or reduced lunch. Measure s Dependent Variables Academic Performance A number of high school academic performance indicators were e mployed as dependent variables and include: exit code GPAs, attendance, cu mulative failed courses b ehavior i ncidents, and s tate exam s cores All three cohort s (2008, 2009, and 2010) had academic data for at least four years of high school. Exit Code U descriptions, the researcher recoded the Exit Codes int o the following categories: 1) transferred out of the district; 2) still in school; 3) d ropped out and did not return; 4) dropped out but returned to graduate; and 5) g raduated DPS reported 3,680 students who graduated, 436 who dropped out, and 70 who ret urned to graduate after having previously dropped out. Due to time limitations, this study will only focus on students who either graduated or dropped out and did not return.

PAGE 28

18 GPAs. Cumulativ e GPA data was provided by year and was measured on a 0 5 scale (w ith 5 given for an A in an Advanced Placement or IB course). E reflects the total GPA for all high school courses completed. For instance, Year 1 includes just 9 th grade; Year 2 includes the cumulative GPAs of Years 1 and 2; etc. Attendance, Cumulative Failed Courses Behavior Incidents, and State Exam Scores. Attendance data is listed as percentage of time attended and includes rates for each year. Cumulative courses failed were reported as one variable to include courses failed in 9 th 12 th g rades Behavior incidents include both expulsions and number of in and out of school suspensions. State exam scores include exam scores for Math, Reading, and Writing for the 9th and 10th grades. These measures were continuous scale scores, which were ver tically equated to allow them to serve as a measure of growth from grade to grade Independent Variables Factor Analysis of Resilience Measures The ScholarCentric assessment includes a total of 108 questions, which are grouped into six resilience factor s and 18 sub scales The researcher wanted to test whether the sub reliability and their predictive power would increase with revised subscale categorizations. The researcher thus ran exploratory factor analysis on the items in each of the original si x factor s. T o do this, t he six factor s were separated into 2 groups: Group A included the items from: Education, Connections, and Stress Group B included the items from: Confidence, Well being, and Motivation The items were analyzed using principal compone nts analysis with varimax rotation as an initial step in examining construct validity by demonstrating that the items are highly correlated with other items in the same subscale, and less correlated with items in other

PAGE 29

19 subscales. Reliability was evaluated using Cronbach alpha estimates of internal consistency. The rotated factor solution identified thirteen factor s, w ith alpha reliabilities rang ing from .679 (one scale) to .93 Six of the new factors were selected for analysis due to their high alpha reliab ilities: Money Difficulties, Ability to Perform in School, Education, Bad Feelings, Intrinsic Motivation, and Family Connections ( see Appendix A for the full survey questions; Appendix B for the factor loadings of the new skill s ; and Appendix C for the rel iability v alues of the new resilience measures ). Independent Variables New Resilience Measures Descriptions Money Difficulties. This measure contains 5 items that assess the degree to which students have difficulty paying for items they need (i.e. food, school supplies). Responses range from 1 ( almost never ) to 5 ( almost always ), with higher scores indicating a greater alpha was .89. 2: Internal Consistency Ability to Perform in School. This measure contains 9 items that assess the degree to which students have difficulty performing academic related tasks (i.e. taking tests, completing homework on time ). Responses range from 1 ( almost never ) to 5 ( almost always ), wi th higher scores indicating a greater degree of difficulty. Internal consistency 83 Education. This measure contains 10 items that assess the degree in which students value education and believ e school and college are important to their future (i.e. it is important that I finish, go to college, etc ). Responses range from 1 ( strongly disagree ) to 5 ( strongly agree ), with higher scores indicating a greater value of education Internal consistency 93

PAGE 30

20 Bad Feelings. This measure contains 14 items that assess how often students have negative health related experiences with manifestations in aggression, physical symptoms, or depressed feelin gs (i.e. breaking things when angry, headaches, feeling hopeless ). Responses range from 1 ( almost never ) to 5 ( almost always ) Note that the company provides to sch which mean that higher scores indicate a higher sense of well being and lower scores indicate a lower sense of well being. Internal consistency coefficient for the total scale alpha was 921 Intrinsic Motivation. This measure contains 7 items that assess whether students go to school because they find it meaningful (i.e. the reason I keep coming to school is because I really enjoy school, because education is important for the goals I have, etc. ). Responses range from 1 ( strongly disagree ) to 5 ( strongly agree ), with higher scores indicating a higher intrinsic motivation Internal consistency coefficient for the total scale using 893 Family Connections. This measure contains 5 items that assess the degree to which students feel supported by their family members (i.e. I am very close to at least one member of my family ). Responses range from 1 ( strongly disagree ) to 5 ( strongly agree ), with higher scores i ndicating stronger support from family Internal consistency 827 Correlations Correlations were run on the new resilience measure s described in section 3.3.b. and the academic performance measur es (3.3.a ). Table 1 reports the means standard

PAGE 31

21 deviations, and correlations among the variables. Cumulative GPA and State Exam Scores generally had the highest correlations of the academic variables. TABLE 1 Summary of Intercorrelations, Means, and Sta ndard Deviations for Scores on SEL Measures and Academic Performance Variables Measure 1 2 3 4 5 6 7 8 9 10 11 12 1. Money Difficulties 2. Ability to Perform in School .577 3. Education .186 .174 4. Bad Feelings .492 .569 .167 5. Intrinsic Motivation .178 .257 .482 .197 6. Family Connections .260 .281 .418 .285 .395 7. Cumulative GPA (Year 4) .227 .213 .206 .169 .163 .137 8. 9 th Grade State Exam Math .211 .232 .168 .170 .087 .095 .635 9. 9 th Grade State Exam Reading .237 .227 .201 .154 .109 .132 .651 .751 10. 9 th Grade State Exam Writing .237 .226 .201 .154 .109 .132 .651 .751 1.000 11. Cumulative Attendance .118 .084 .111 .108 .078 .037 .5 31 .252 .234 .234 12. Cumulative Courses Fail ed .151 .150 .129 .132 .111 .075 .720 .397 .378 .377 .507 M 1.72 2.30 4.52 4.02 3.94 4.35 2.85 5 5 6. 66 550.09 5 49.98 .85 4. 8 6 SD .84 .83 .59 .81 .80 .70 .96 67.8 8 70.49 70.47 .14 6.69 *Correl ation is sig nificant at the 0.05 level (2 tailed). All other va lues are significant at the 0.01 level (2 tailed). Procedures of Data C ollection Academic Performance Data Academic performance data was provided by DPS and includes the following variables: GPAs, beha vior incidents, cumulative courses failed attendance, state exam scores, and graduation vs dropout status This data was provided for up to five years of

PAGE 32

22 grades nine, ten, eleven, twelve, and fifth year twelfth. D ue to time limitations, this study will only include data for grades nine through twelve, decision made based on research showing that the majority of students who graduate do so on time (Neild & Balfanz, 2006). The researcher matched the resi lience and a cademic data by district assigned Student IDs and then stripped the file of the ids to maximize the protection of student privacy. To account for students who repeated one or more grades, the researcher categorized the data by year, rather than by grade. T able 2 illustrates this categorization system. Eighth grade year indicates the year the student ended the 8 th grade (i.e. Cohort 1 was in 8 th grade during the 2007 2008 school year), and Year 1 indicates the year the student first entered high school. Yea rs 2 5 indicate the subsequent years the student was enrolled in high school. TABLE 2 Data Categorization by Year 8 th grade Y ear Year 1 Year 2 Year 3 Year 4 Year 5 Cohort 1 2008 2009 2010 2011 2012 2013 Cohort 2 2009 2010 2011 2012 2013 2014 Cohort 3 2010 2011 2012 2013 2014 2015 Resilience Data resilience assessment. The survey was administered to students the summer prior to entering the ninth grade. Teachers and other ed ucators administered the assessment as part of normal education practice in a normal education setting. At no time did the researcher have contact with students. The ScholarCentric resilience survey consists of

PAGE 33

23 108 questions is available in English and Sp anish, and takes students approximately 30 minutes to complete ( See Appendix A for the full survey questions). Missing D ata. Missing values were not imputed for dependent variables; cases with missing data for a dependent variable were dropped from the a nalysis involving that dependent variable. Missing items within the social emotional subscales ranged from .4% to 4.7% of cases, depending upon the subscale. In those cases, if two or more items were answered, the mean of the answered items was substituted for missing items within the subscale. If fewer than two of the items for a scale were complete, the case was dropped from the analysis.

PAGE 34

24 CHAPTER IV RESULTS The analy sis employed linear regression and logistic regression to determine the percent of vari ance in GPA, cumulative courses failed, and attendance attributable to the measured social and emotional variables. Due to time and data availability constraints, the academic measures were limited to these variables. Bivariate correlations among the inde pendent variables ( see Table 1 ) were examined to identify variables with the highest correlations that would contribute to multicollinearity if used together in a regression analysis. The three variables Money Difficulties, Importance of Education, and Ba d Feelings emerged as the high est social and emotional bivariate correlates of Year 1 Attendance (r= .133, r=.145, and r=.159, respectively). The remaining independent variables were omitted to avoid multicollinearity. In a linear regression analysis, thes e variables explained 4.2% of the variance in Year 1 Attendance (R=.205, R2=.042, p < .001). Residual attendance values, in which the variance explained by the resilience measures was removed, were saved. Linear regression found that residual attendance ex plained 31% of the variance in Year 1 GPA (R=.553, R2=.306, p < .000) and 18% of the variance in cumulative courses failed (R=.419, R2=.175, p < .000). Multiple regression was then used to account for variance in GPA associated with resiliency after contro lling for attendance. The three resiliency variables (Money Difficulties Importance of Education, and Bad Feelings) explained 12.4% of the variance of the residual GPA values (R=.353, R 2 =.124, p < .000) and 4.4% of the variance of the residual cumulative courses failed (R=.211, R 2 =.044, p < .000).

PAGE 35

25 Logistic regression found that Year 1 Attendance and Year 1 GPA correctly predicted 33.9% of students who dropped out and 98.2% of students who graduated (cut value=.5, Cox & Snell R2=.196, Nagelkerke R2=.404 ). These academic measures combined with the three resilience measures correctly predicted 33.1% of students who dropped out and 98.3% of students who graduated (cut value=.5, Cox & Snell R2=.192, Nagelkerke R2=.404). When the cut value was increased to .9 (to err on the side of over identifying students predicted to drop out), logistic regression found that the academic measures correctly predicted 79.7% of students who dropped out and 80.8% of students who graduated (Cox & Snell R2=.196, Nagelkerke R2=.40 4). Using this same cut value, t he academic measures combined with the resilience measures correctly predicted 79.1% of students who dropped out and 81.6% of students who graduated (Cox & Snell R2=.192, Nagelkerke R2=.404). The inclusion of the three resil iency variables did not increase the accuracy of prediction of dropouts or graduates. resilience measures accurately identified dropouts that were not correctly identified by t he academic measures. The resilience measures correctly identified 8.1% of students who dropped out and had been incorrectly identified as predicted to graduate by the academic measures Both the academic and resilience measures incorrectly identified 12% of dropouts.

PAGE 36

26 CHAPTER V DISCUSSION AND CONCLUSION their graduation rates, and they depend on readily accessible useful/actionable data and effective EWIs that accurately predi evaluated 1) whether 8th grade resilience measures are predictive of various high school academic outcomes, including dropout rates; 2) the relationship between academic measures and social and emotional measures and their ability to predict graduation vs. dropout; and 3) whether the resilience measures are able to effectively predict dropouts who are not identified by ABC measures. T he three social and emotional measures Money Difficulties, Importance of Education, and Bad Feelings were found to be significant predictors of Year 1 GPA and cumulative courses failed, after controlling for the amount of variance explained by these resilience measures on attendance. Attendance was selected as the academic ben chmark success ( Bruce, et al. 2011; Pinkus, 2008 ). Attendance is so important because this measure is highly correlated with a variety of other academic variables inclu ding GPAs, state exam scores, fail rates, dropping out, etc. ( Balfanz, Herzog, & MacIver, 2007 ). In short, a missed day of school is a missed opportunity to learn. Behavior incidents were not included in the regression analysis due to a large amount of mi ssing data on this variable. Due to time constraints, the researcher focused on GPA, cumulative failed courses, attendance and graduation vs. dropout status. It would be interesting to revisit the analysis and include state exam scores, since these had

PAGE 37

27 som e of the highest correlat ions of the academic variables with the resilience measures. GPA and attendance alone correctly predicted 79.7% of students who dropped out and 80.8% of students who graduated ; these findings are consistent with earlier research me ntioned in the Literature Review that established the ABC indicators as predictive of future academic outcomes ( Bruce, et al., 2011; Pinkus, 2008 ). This amount of predictive power was retained when the three resilience measures were combined with GPA and a ttendance: the combined measures correctly predicted 79.1% of students who dropped out and 81.6% of students who graduated. These results were dependent on a cut value of .9; t he researcher made the decision to err on the side of over predicting dropouts b ecause the risks of false positives outweigh the risk of under identifying students at risk of dropping out. T the researcher discovered that the resilience measures correctly predicted 8.1% of students who dropped out that were incorrectly predicted to graduate by the academic measures This suggest s that the resilience measures add additional predictive value in that they are able to identify some students who are at risk of dropping out but f e ll under the radar of the academic measures. A total of 12% of students who dropped out were not correctly identified by either the academic or resilience measures. This information reveals that there are other elements beyond academic and social and em otional factors that are contributing to student dropouts. In addition to the ability to effectively predict academic performance and graduation vs. dropout status, resilience measures provide educators with information on disengagement and likelihood of dropping out. This can

PAGE 38

28 individualized needs and create customized interventions accordingly. Further, because resilience is malleable, e ducators can improve academic outcomes through SEL instruction. Resilient students are more self regulated and engaged learners, and they perceive the value of education on their future goals (Zimmerman, 2011; Hulleman et al., 2010; Hulleman & Harackiewicz 2009). The reader should be careful to not generalize the results of this correlational study. The sample of this research was limited to a single school district, and replication with other school districts would strengthen the analysis. In addition, s ocial and emotional measures were assessed using self report data, and therefore all limitations associated with traditional survey research are warranted. Finally, the analysis may be strengthened by an analysis of structural equation modeling (SEM) and/ or path analysis A main benefit of SEM (an extension of multiple regression) is that it is able to examine relationships between latent constructs represented by multiple measures and isolate observational error from measurement of latent variables ( Lei & Wu, 2007 ) Path analysis ( which falls under the umbrella of SEM) may offer more explanatory power than the methods used in this project because it is able to accommodate the existence of multiple dependent variables. Path analysis identifies three types o f effect: 1) direct effects; 2) indire c t effects; and 3) t otal effect. This method may better explain the complex and intertwined relationships of the factors contributing to student dropouts because it accounts for the association of one variable with ano ther mediated through other variables in the model ( Streiner, 2005) warning systems meet the criteria required for an effective early warning system

PAGE 39

29 (research and evidence bas ed; readily accessible data points; and effective predictor of gradua tion vs. dropout status) and may result in a more effective system for a number of reasons. First, social and emotional factors, combined with GPA and attendance are fairly strong predict status. Second, by understanding why students are struggling academically educators can This stud y focused on the resilience measures Importance of Education, Bad Feelings, and Money Difficulties. sphere of influence, schools can and should provide resources to help students cope with bad feelings, and schools need to work on providing motivating environments and persuading students that education has intrinsic value and is not simply a societal expectation. And thirdly, these factors may be able to accurately identify a percentage of students at risk for dropping out of high school who are not marked at risk by academic measures.

PAGE 40

30 REFERENCES Allensworth, E. & Easton, J. (2007). What matters for staying on track and graduating in Chicago Public High Schools. Chicago: Consortium on Chicago School Research. Alliance for Excellent Education. (2011). The high cost of high school dropouts: What the nation pays for inadequate high schools. Washington D.C.: MetLife Foundation Retrieved from ERIC database. (ED 537542) Balfanz, R. H erzog, L. MacIver, D. (2007). Preventing student disengagement and keeping students on the graduation path in urban middle grades schools: Early identification and effective interventions. Educational Psychologist 42 (4), 223 235. Balfanz, R. (2009). Put ting middle grade s tudents on the graduation path: A policy and practice b rief. Westerville, OH: National Middle School Association Balfanz, R., Fox, J. H., Bridgeland, J.M., & McNaught, M. (2009). Grad Nation: A guidebook to help communities tackle the dropout crisis. Washington, D.C.: America's Promise Alliance, Civic Enterprises, & Everyone Graduates Center. Retrieved from ERIC database. (ED 505363) Balfanz, R., Bridgeland, J., Fox, J.H., DePaoli, J., Ingram, E., & Maushard, M. (2014). Building a Gra dNation: Progress and challenge in ending the high school dropout epidemic 2014 Annual Update Washington, D.C.: America's Promise Alliance, Alliance for Excellent Education, Civic Enterprises, & Everyone Graduates Center. Retrieved from http://gradnation.org/resource/building gradnation progress and challenge ending high school dropout epidemic 2014 Baltes, P. B. (1997). O n the incomplete architecture of human ontogeny: Selection, optimization, and compensation as foundation of developmental theory. American Psychologist 52 366 380. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory Englewood Cliffs, NJ: Prentice Hall. rhetoric: Creating healthy communities for children and adolescents. Applied Developmental Science, 2 (3), 138 159. Bernard, B. (20 04 ). Resiliency: What we have learned San Francisco: WestEd. Blum, R.W., & Libbey, H.P. (2004). School connectedness Strengthening health and education outcomes for teenagers. Journal of School Health, 74, 229 299.

PAGE 41

31 Bruce, M., Bridgeland, J. M., Fox, J. H., & Balfanz, R. (2011). On track for success: The use of early warning indicator and intervention systems to build a grad nation. Baltimore, MD: Johns Hopkins University, School of Education, Everyone Graduates Center. Retrieved from ERIC database. (ED 526421) Burd Sharps, S. & Lewis, K. (2012). One in seven: Ranking youth disconnection in the 25 largest metro areas. Measure of America of the Social Science Research Council. Retrieved from http://ssrc static.s3.amazonaws.com/moa/MOA One_in_Seven09 14.pdf Cheesman Day, J., & Newburger, E. C. (2002). The big payoff: Educational attainment and synthetic estimates of work life earnings. Current Population Reports (pp. 23 210 ). Washington, DC: U.S. Census Bureau. Retrieved from https://www.census.gov/prod/2002pubs/p23 210.pdf Close, W., & Solberg, S. (2008). Predicting achievement, distress, and retention amon g lower income Latino youth. Journal of Vocational Behavior, 72 31 42. doi: 10.1016/j.jvb.2007.08.007 Cohen, S., & Wills, T. A. (1985). Stress, social suppor t and the buffering hypothesis. Psychological Bulletin 98, 310 357. doi: 10.1037//0033 2909.98.2.310 Cotton, K. (2003). Principals and student achievement: What the research says Alexandria, VA: Association for Supervision and Curriculum Development. Cutler D & Lleras Muney A. (2006) Education and health: Evaluating t heories and evidence. Bethesda, MD: National Bureau of Economic Research, 2006. Data Quality Campaign. (2013). Supporting early warning systems: Using data to keep students on track to success. Retrieved from: http://www.dataqualitycampaign.org/files/Supporting%20Early%20Warning%20 Systems.pdf Davis, A., Solberg, V.S., De Baca, C., & Hargrove, T. (2014). Use of social em otional learning skills to predict future academic success and progress toward graduation. Journal of Education for Students Placed at Risk (JESPAR), 19 :3 4, 169 182, DOI 10.1080/10824669.2014.972506. Deci, E. L., Schwartz, A. J., Sheinman, L., & Ryan, R M. (1981). An instrument to assess adults' orientations toward control versus autonomy with children: Reflections on intrinsic motivation and perceived competence. Journal of Educational Psychology, 73, 642 650. doi: 10.1037//0022 0663.73.5.642 Dohrenwend, B. S., & Dohrenwend, B. P. (Eds.). (1974). Stres sful life events: Their natur and effects New York: Wiley.

PAGE 42

32 Dornbusch, S. M., Mont Reynaud, R., Ritter, P. L., Chen, Z., & Steinberg, L. (1991). Stressful events and their correlates among adolescents of diverse backgrounds. In Colton, M. E., & Gore, S. ( E ds.), Adolescent stress: Causes and consequences (pp. 111 130) New York,, NY: Aldine de Gruyter Durlak, J. A., Weissberg, R. P., Dym nicki, A. B., Taylor, R. D. & Schellinger, K. B. analysis of school based universal interventions. Child Development, 82 (1): 405 432. Eccles, J. S. (2005). Subjective task va lue and the Eccles et al. model of achievement related choices. In A. J. Elliot & C. S. Dweck (Eds.), Handbook of competence and motivation (pp. 105 121). New York, NY: Guilford. Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goal s. Annual Review of Psychology 53 (1), 109 132. doi:10.1146/annurev.psych.53.100901.135153 Egeland, B., Carlson, E., & Sroufe, L. A. (1993). Resilience as process. Development and Psychopathology 5 517 528. doi: 10.1017/s0954579400006131 Evans, G.W. (2004). The environment of childhood poverty. American Psychologist 59 (2), 77 92. Georgetown University Center on Education and the Workforce (2010). Help wanted: Projections of jobs and educa tion requirements through 2018. Retrieved from www9.georgetown.edu/grad/gppi/hpi/cew/pdfs/FullReport.pdf Gillis, S. A. (2011). A structural equation model of resiliency in adolescence ( Unpu blished doctoral dissertation ). Un iversity of Wisconsin Milwaukee, Milwaukee, Wisconsin. Gillock, K. L., & Reyes, O. (1999). Stress, supports, and academic performance of urban, low income, Mexican American adolescents. Journal of Youth and Adolescence 28 (2) 259 282. doi: 10.1023/a:1021657516275 Grossman M., & Kaestner R. (1997). Effects of education on health. In: The Social Benefits of Education J.R. Berman & N. Stacy (E ds). Ann Arbor, MI: Univ ersity of Michigan Press. H anson, T.L., & Austin, G. (2003). Student health risks, resilience, and academic per formance in California: Year 2 report, longitudinal analyses Los Alamitos, CA: WestEd.

PAGE 43

33 Harackiewicz, J. M., Durik, A. M., Barron, K. E., Linnenbrink Garcia, L., & Tauer, J. (2008). The role of achievement goals in the development of interest: Reciprocal relations between achievement goals, interest, and performance. Journal of Educational Psychology 100 105 122. doi: 10.1037/0022 0663.100.1.105 Hobfoll, S. (1989). Conservation of resources: A new attempt at conceptualizing stress. American Psychologist 44 (3), 513 524. doi:10.1037/0003 066X.44.3.513. Hobfoll, S. E. (1998). St ress, culture, and community, the psychology and philosophy of stress. Dordrecht, the Netherlands: Kluwer Academic/Plenum Publishers. Hulleman, C. S., Godes, O., Hendricks, B. L., & Harackiewicz, J. M. (2010). Enhancing interest and performance with a uti lity value intervention. Journal of Educational Psychology 102 (4), 880 895. doi: 10.1037/a0019506 Hulleman, C. S., & Harackiewicz, J. M. (2009). Promoting interest and performance in high school science classes. Science 326 1410 1412. doi: 10.1126/science.1177067 Jobs for the Future. (2014). Early warning indicators and segmentation analysis: A technical guide on data studies that inform dropo ut prevention and recovery. Boston, MA.: Jobs for the Future. Klem, A.M., & Connell, J.P. (2004). Relationships matter: Linking teacher support to student engagement and achievement. Journal of School Health, 74 262 273; doi :10.1111/j.1746 1561.2004.tb 08283.x. Lei, P. & Wu, Q. (2007). Introduction to structural equation modeling: Issues and practical considerations. Educational Measurement: Issues and Practices, 26 (3), 33 43. Retrieved from http://ncme.org/linkservid/47EFEB5A 1320 5 CAE 6EC90BFDF09AA39E /showMeta/0 Lewis, K. M., Byrd, D. A., & Ollendick, T. H. (2012). Anxiety symptoms in African American and Caucasian youth: Relations to negative life events, social support, and coping. Journal o f Anxiety Disorders 26 (1), 32 39. doi:10.1016/j.janxdis.20 11.08.007 Lezotte, L. W., & McKee Snyder, K. (2011). What effective schools do: Re envisioning the correlates. Bloomington, IN: Solution Tree Press. Low M D Low B J Baumler E R ., & Huynh, P.T. (2005). Can education policy be health policy? Implica tions of research on the social d eter minants of health. J ournal of Health Polit ics, Policy and Law, 30 (6): 1131 62, 2005.

PAGE 44

34 McEwen, B. S. (1998). Protective and damaging effects of stress mediators. New England Journal of Medicine, 322 171 179. Miller, W., Simon, P., & Maleque, S. (2009). Beyond health care: New directions to a healthier America. Washington, D.C.: The Robert Wood Johnson Foundation Commission to Build a Healthier America. Mirowsky J ., & Ross C E. (2003). Education, social status, and health. H awthorne, NY: Aldine de Gruyter Multon, K. D., Brown, S. D., & Lent, R. W. (1991). Relation of self efficacy beliefs to academic outcomes: A meta analytic investigation. Journal of Counseling Psychology, 38 30 38. doi: 10.1037//0022 0167.38.1.30 Neild, R. C., & Balfanz, R. (2006). Unfulfilled promise: The dimensions and 2005. Baltimore, MD: Center for Social Organization of Scho ols, Johns Hopkins University. Neild, R. C., and Balfanz, R. & Herzog, L. (2007). An early warning system. Educational Leadership 65 .2 (2007): 28 33. Pinkus, L. (2008). Using early warning data to improve graduation rates: Closing cracks in the educatio n system. Washington, DC: Alliance for Excellent Education. Reyes, O. & Jason, L.A. (1993). Pilot study examining fa ctors associated with academic success for Hispanic high school students. Journal of Youth and Adolescence 22 57 71. Ross C E ., & Miro wsky J. (1999). Refining the association between education and h ealth: The e ff ects of quantity, credential, and selectivity. Demography, 36 (4): 445 60, 1999. Ryan, R. M., & Deci, E. L. (2002). An overview of self determination theory. In E. L. Deci & R. M. Ryan (Eds.), Handbook of self determination research (pp. 3 33). Rochester, NY: University of Rochester Press. Ryan, R. M., & Deci, E. L. (2008). A self determination approach to psychotherapy: The motivational basis for effective change. Canadian Psyc hology 49 186 193. doi: 10.1037/a0012753 Scales, P. C., Benson, P. L., Leffert, N., & Blyth, D. A. (2000). Contribution of developmental assets to the prediction of thriving among adolescents. Applied Developmental Science, 4 (1), 27 46.

PAGE 45

35 Scales, P. C., Roehlkepartain, E.C., Neal, M., Kielsmeier, J.C., & Benson, P.L. (2006). The role of developmental assets in predicting academic achievement: A longitudinal study. Journal of Adolescence, 29 (5 ), 692 708. Washington, DC: Robert Wood Johnson Foundation Commission to Build a Healthier America, 2009. Schunk, D. H., & Pajares, F. (2009). Self efficacy theory. In K. R. Wentzel & A. Wigfield (Eds.), Handbook of motivation at school (pp. 35 53). New York NY : Ro utledge. Solberg, V.S.H., Gusavac, N., Hamann, T., Felch, J., Johnson, J., Lamborn, S., & Torres, J. (1998). The Adaptive Success Identity Plan (ASIP): A career intervention for college students. Career Developments Quarterly 47 48 95 Solberg, V. S., R., & Davis, B. ( 1993 ). Self efficacy and Hispanic college students: Validation of the College Self Efficacy Instrument Hispanic Journal of Behavioral Sciences 15 80 95. doi: 10.1177/07399863930151004 Solberg, V. S. H., & Villareal, P. (1997). Self efficacy, stress, and social support as predictors of Hispanic college adjustment. Hispanic Journal of Behavioral Sciences,19 182 201. Solberg, S., Carlstrom, A., Howard, K., & Jones, J. (2007). Classifying at risk high school youth: The influence of community violence and protective factors on academic and health outcomes, Career Development Quarterly 55 313 327. Stetser, M., & Stillwell, R. (2014). Public h igh school four year on time graduation rates and event dropout rates: School years 2010 11 and 2011 12. First Look (NCES 2014 391). Streiner, D.L. (2005). Finding our way: An introduction to Path Analysis. Canadian Journal of Psychiatry, 50 (2), 115 12 2. Retrieved from http://www.unt.edu/rss/class/mike/6810/articles/pathintro.pdf Suldo, S. M., & Thalji, A., & Ferron, J. (2011). Longitudinal academic outcomes pred being, psychopathology, and mental health status yielded from a dual factor model. Journal of Positive Psychology, 6 (1), 17 30 doi: 10.1080/17439760.2010.536774 Toldson, I.A. (2008). Breaking b arriers: Plottin g the path to academic success for school age African American m ales. Washington, D.C.: Congressional Black Caucus Foundation, Inc. Torres, J.B., & Solberg, V. S. H. (2001). Role of self efficacy, stress, social integration, and family support in Latino c ollege student persistence and health. Journal of Vocational Behavior, 59 53 63.

PAGE 46

36 U.S. Department of Education, National Center for Education Statistics, Common Core of Data (CCD). (2014). Public high school four year on time graduation rates and event dro pout rates: School years 2010 11 and 2011 12. (NCES 2014 391) Retrieved from School Year 2011 12, Preliminary Version 1a. Retrieved from http://nces.ed.gov/programs/coe/indicator_coi.asp U.S. Department of Education, National Center for Education Statistics. (1998 2011). Public elementary/secondary school universe surveys Retrieved from https://nces.ed.gov/ccd/pdf/psu98rgen.pdf Valdez, C. R., Lambert, S. F., & Ialongo, N. S. (2011). I dentifying patterns of early risk for mental h ealth and academic problems in adolescence: A longitudinal study of urban youth. Child Psychiatry and Human Development 42 521 538. doi: 10.1007/s105 78 011 0230 9 Waxman, H.C. & Huang, S.L. (1997). Classroom instruction and learning environment differences between effective and ineffective urban elementary schools for African American students. Urban Education, 32 (1), 7 44. Windle, M., & Windle, R. C. (1996). Coping strategies, drinking motives, and stressful life events among middle adolescents: Associations with emotional and behavioral problems and with academic functioning. Journal of Abnormal Psycholology, 105 (4) 551 560. doi: 10.1037//0021 843x.105.4.551 Werner, E., & Smith. R. (1992). Overcoming the o dds: High r isk c hildren from b irth to a dulthood. New York: Cornell University Press. What is social and emotional learning? (2 015). In CASEL. Retrieved May 3, 2015, from http://www.casel.org/social and emotional learning Winkleby M A Fortmann S P ., & Barrett D C (1990). Social c lass disparities in risk fa ctors for disease: Eight year p revalence p atterns by level of education. Prev Med 19 (1): 1 12, 1990. Zimmerman, B. J. (2011). Motivational sources and outcomes of self regulated learning and performance. In B. J. Zimmerman & D. H. Schunk (Eds.), Handbook of self regulation of learning and performance (pp. 49 64). New York, NY: Routledge.

PAGE 47

37 APPENDIX A Importance of Education This section asks about your beliefs about the importance of school and college. Mar k the number on the answer sheet that best represents your present attitude of opinion. Remember, this is not a test, and there are no right or wrong answers. The range of answers is: 1 = Strongly disagree 2 = Disagree 3 = Neutral/undecided 4 = Agree 5 = S trongly agree Using the scale above, please mark the number on the answer sheet that best shows *the degree to which you agree with each statement below:* 1. Finish school. 2. Do well in school. 3. Go to college. 4. Do well in college. 5. Make sure my teac her knows that I want to do well in school. 6. Find out about colleges. 7. Learn how to be successful in college. 8. Get good grades in school. 9. Learn how to be successful in school.

PAGE 48

38 10. Get a college degree. Confidence This section asks for information about the degree of confidence you have in completing a variety of activities with being a student at your school. Mark the number on the answer sheet that best represents your present attitude or opinion. Remember, this is not a test, and there are no ri ght or wrong answers. The range of answers is: 1 = Not confident at all 2 = Mostly not confident 3 = Somewhat confident 4 = Mostly confident 5 = Extremely confident Using the scale above, please mark the number on the answer sheet that best shows *the degr 11. Making new friends at school. 12. Talking to teachers about homework. 13. Taking good notes in class. 14. Writing a paper for English class. 15. Joining a sports activity. 16. Understanding what you read in your schoolbooks. 17. Asking a question in class. 18. Joining an after0school club. 19. Correctly figuring out math problems.

PAGE 49

39 20. Turning in your assignments on time. 21. Going to class every day. 22. Working on a group class project. 23. Getting along with classmates. 24. Doing well on your tests. 25. Using a computer to write a paper. 26. Using the library. 27. Using a computer to write a paper. 28. Participating in class discussions. 29. Keeping up to date on schoolwork. 30. Preparing for a test. 31. Relaxing during a test. 32. Studying with others for a test. Connections This section asks your relationships with family, teachers, and friends. Mark the number on the answer sheet that best represents your present attitude or opinion. Remember, this is not a test, and there are no right or wrong answers. The range of answers is: 1 = Strongly disagree 2 = Disagree 3 = Neutral/undecided 4 = Agree 5 = Strongly agree

PAGE 50

40 Please indicate, by marking the number on the answer sheet that best represents the dgree t o which you agree with the following statements: 33. There is a family member who I can talk to about important decisions in my life. 34. Members of my family recognize my abilities and skills. 35. There is no one in my family who shares my interests and concerns. 36. I am very close with at least one other member of my family. 37. There is no on in my family with whom I feel comfortable talking about my problems. 38. I can talk about school issues of concerns with a family member. 39. There are family mem bers I can count on in an emergency. 40. Teachers here care about their students. 41. There is a teacher here I can go see to talk about academic problems. 42. Teachers here respect me. 43. Teachers here are interested in my success. 44. There is a teacher here I can talk to about a personal problem. 45. I have friends here at school. 46. There are friends I can talk to about important decisions. 47. There is a friend I can depend on for help. 48. I have no friends I can depend on.

PAGE 51

41 Stress This section ask s about the stresses in your life. Mark the number on the answer sheet that best represents your present attitude or opinion. Remember, this is not a test, and there are no right or wrong answers. The range of answers is: 1 = Almost never 2 = Not very ofte n 3 = Somewhat often 4 = Very often 5 = Almost always Please indicate the degree to which you have experienced each of the following in the PAST MONTH: 49. Difficulty trying to fulfill responsibilities at home and at school. 50. Difficulty trying to meet friends. 51. Difficulty taking tests. 52. Difficulty talking with teachers about schoolwork. 53. A fear of failing to meet family expectations. 54. Difficulty asking questions in class. 55. Difficulty living in the local community. 56. Difficulty understan ding how to use the school library. 57. Difficulty handling relationships. 58. Difficulty handling your schoolwork load. 59. Difficulty with classmates treating you differently than they treat each other. 60. Difficulty writing papers for class.

PAGE 52

42 61. Diffic ulty learning how to use computers. 62. Difficulty paying for school supplies. 63. Money difficulties due to owing money to others. 64. Difficulty paying for food. 65. Difficulty paying for recreation and entertainment. 66. Difficulty due to your family ex periencing money problems. 67. Difficulty getting your homework done on time. 68. Difficulty because of feeling a need to perform well in school. 69. Difficulty from teachers. 70. Difficulty from classmates. Well b eing This section asks you about how ofte n you have had any of these health related experiences during the past week. Mark the number on the answer sheet that best represents your present attitude or opinion. Remember, this is not a test, and there are no right or wrong answers. The range of answ ers is: 1 = Almost never 2 = Not very often 3 = Somewhat often 4 = Very often 5 = Almost always Please indicate the degree to which you have experiences each of these during the PAST WEEK:

PAGE 53

43 71. Being tried but unable to slee p. 72. Mood swings. 73. Feelings of danger. 74. Feeling depressed. 75. Feelings of self doubt. 76. Nightmares. 77. Snacking more than usual. 78. Feeling hopelessness. 79. Sleeping less than usual at night. 80. Getting sick a lot. 81. Overeating. 82. Break ing things when angry. 83. Headaches. 84. Increased heartbeat. 85. Fighting with friends. 87. Losing your temper. 89. Not sleeping well. 90. An upset stomach. 91. Inability to sleep. 92. Increased appetite.

PAGE 54

44 93. B ecoming easily upset. Motivation This section asks about your reasons for going to school. Different people have different reasons for going to school; we just want to know how much you agree or disagree with each reason given below. Mark the number on t he answer sheet that best represents your present attitude or opinion. Remember, this is not a test, and there are no right or wrong answers. The range of answers is: 1 = Strongly disagree 2 = Disagree 3 = Unsure/undecided 4 = Agree 5 = Strongly agree The 94. Because I really enjoy school. 96. So I can make lots and lots of money. 97. Because education is important for the goals I have. isappointed in me. 102. Because skills like reading, math, and science are important to me.

PAGE 55

45 104. Because failing to get my diploma would bother and disappoint me. 105. Because there are a lot of interesting things to do. 106. Because I see the importance of learning. 107. Because, to me, education is important. about it.

PAGE 56

46 APPENDIX B F ACTOR LOADINGS OF SIX NEW RESILIENCE MEASURES CONSIDERED FOR THE MULTIPLE REGRESSION IN THIS STUDY (Based on Principal Components Analysis with Varimax Rotation) Component Matrix Component 1 2 3 4 5 6 1. Finish school. .751 2. Do well in school. .788 3. Go to college. .811 4. Do well in college. .828 5. Make sure my teacher knows that I want to do well in school. .664 6. Find out about colleges. .723 7. Learn how to be successful in college. .827 8. Get good grades in school. .770 9. Learn how to be successful in school. .757 10. Get a college degree. .806 33. There is a family member who I can talk to about important decisions in my life. .781 34. Members of my f amily recognize my abilities and skills. .693 36. I am very close with at least one other member of my family. .722 38. I can talk about school issues of concerns with a family member. .764 39. There are family members I can count on in a n emergency. .709 49. Difficulty trying to fulfill responsibilities at home and at school. .690 51. Difficulty taking tests. .718 52. Difficulty talking with teachers about schoolwork. .706 53. A fear of failing to meet family expec tations. .636 54. Difficulty asking questions in class. .673 58. Difficulty handling your schoolwork load. .792 60. Difficulty writing papers for class. .714 67. Difficulty getting your homework done on time. .756 68. Difficul ty because of feeling a need to perform well in school. .739

PAGE 57

47 Component Matrix (Continued) Component 1 2 3 4 5 6 62. Difficulty paying for school supplies. .562 .603 63. Money difficulties due to owing money to others. .565 .538 64. D ifficulty paying for food. .573 .625 65. Difficulty paying for recreation and entertainment. .594 .616 66. Difficulty due to your family experiencing money problems. .598 .558 72. Mood swings. .668 73. Feelings of danger. .658 74. Feeling depressed. .729 75. Feelings of self doubt. .710 78. Feeling hopelessness. .726 82. Breaking things when angry. .643 83. Headaches. .558 84. Increased heartbeat. .645 85. Fighting with friends. .636 86 .703 87. Losing your temper. .694 .577 90. An upset stomach .643 93. Becoming easily upset. .707 94. Because I really enjoy school. .669 97. Because education is important for th e goals I have. .698 .647 102. Because skills like reading, math, and science are important to me. .706 105. Because there are a lot of interesting things to do. .727 106. Because I see the importance of learn ing. .789 107. Because, to me, education is important. .780 Factor loadings < .35 are suppressed.

PAGE 58

48 APPENDIX C N EW SUB SKILLS, QUESTION NUMBERS, AND RELIABILITY VALUES New Sub Scale Items Reliability Money Difficulties questions 62, 63, 64, 6 5, and 66 0.89 0 Ability to Perform in School questions 49, 51, 52, 53, 54, 58, 60, 67, and 68 0.883 Education questions 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 0.93 0 Teacher Connections questions 40, 41, 42, 43, and 44 0.821 Family Connections questions 33, 34, 36, 38, and 39 0.827 Family Connections group 2 questions 35 and 37 0.679 Peer Connections questions 45, 46, and 47 0.814 Bad Feelings questions 72, 73, 74, 75, 78, 82, 83, 84, 85, 86, 87, 88, 90, and 93 0.921 Academic Confidence questions 13, 16, 19, 20, 24, 29, and 30 0.864 Intrinsic Motivation questions 94, 97, 99, 102, 105, 106, and 107 0.893 Extrinsic Motivation questions 98, 100, 101, and 103 0.71 0 Sleeping Problems questions 71, 79, 89, and 91 0.89 0 Eating Problems questions 77 81, and 92 0.806