learning environment and its impact on student’s academic achievement

irakan_0

ed521456.pdf

Home >English homework help >learning environment and its impact on student’s academic achievement

Educational Resilience: The Relationship Between School

Protective Factors and Student Achievement

Eric J. Banatao

A dissertation submitted to the faculty of

San Diego State University

In partial fulfillment of the requirements for the degree

Doctor of Educational Leadership

May 10, 2011

SAN DIEGO STATE UNIVERSITY

The Undersigned Faculty Committee Approves the

Dissertation of

Dr. Eric J. Banatao

Educational Resilience: The Relationship Between School

Protective Factors and Student Achievement

_____________________________________________ Cynthia L. Uline, Chair

College of Education, Educational Leadership

_____________________________________________ Margaret R. Basom

College of Education, Educational Leadership

_____________________________________________ Maruta R. Gardner

Educational Consultant

______________________________ Approval Date

iii

Eric J. Banatao

DEDICATION

To my life partner, best friend, and most beautiful spirit I know, Andrea Most.

I am in you and you in me, mutual in divine love.

—William Blake (1757-1827)

ABSTRACT

Educators are increasingly pressured to raise standardized test scores under the No

Child Left Behind (NCLB) Act of 2001, which has resulted in increased instructional

time in tested subjects and test-focused school leaders who neglect school climate factors

which have been associated with positive student development and increased student

achievement. The theoretical framework of resilience, applied to the school setting, along

with associated school climate data, may offer keys to improved school organization,

instructional delivery, data analysis, and teacher training, resulting in improved student

outcomes. The California Healthy Kids Survey (CHKS) and its Resilience Youth

Development Module (RYDM) represent a research-based, psychometrically-sound

instrument that measures school climate elements, such as external school protective

factors, internal student assets, and school connectedness.

The independent variables of this study included external school protective

factors, such as: caring adults, high expectations, and opportunities for meaningful

participation; internal student assets, such as: problem-solving, self-efficacy, empathy,

and self-awareness; demographic control variables, such as percent number of students:

African-American, Hispanic/Latino, participating in free/reduced meals, and English

language learners; and a school connectedness variable. Aggregated school-level scores

were drawn from 1.5 million student cases (n = 1143, 987, and 836 schools in 2004,

2006, and 2008, respectively). The dependent variables were school Academic

Performance Index (API) scores. This study investigated the relationship between select-

CHKS items and subscales to a student achievement measure; school API score, a figure

calculated by California Department of Education’s general accountability system based

on standardized test performance. This correlational study with replicated procedures

across three sets of data examined matching 7th grade CHKS data and school API scores

through descriptive and inferential statistical analyses in school years 2003-2004, 2005-

2006, and 2007-2008. A three-part statistical procedure for data analysis included a

zero-ordered simple correlation to school API, then two forced-entry hierarchical multiple

regression analyses that accounted for the effects of all variables, and the tested effect of

the mediator variable, school connectedness.

Study findings indicated that the school meaningful participation and school

connectedness variables demonstrated statistically significant positive correlations to

school API scores through three study replications, after accounting for the effect of all

other study variables, such that the higher the reports of school meaningful participation

and school connectedness, the higher the school API score. School connectedness,

however, was three to four times a more powerful predictor of school API scores than

school meaningful participation. The study findings support educational leadership

approaches and policy development efforts that purposefully bolster school

connectedness and school meaningful participation to more positively impact student

learning and school reform efforts.

vii

TABLE OF CONTENTS

PAGE

ABSTRACT.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

LIST OF TABLES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

LIST OF FIGURES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiv

ACKNOWLEDGMENTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

CHAPTER 1—INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Statement of the Problem.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

Purpose of the Study. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Research Questions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

Summary of the Literature. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

Method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

Sample. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Research Design.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Variables of Interest. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

Data Analysis Procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Research Significance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

Definition of Terms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

CHAPTER 2—REVIEW OF THE LITERATURE. . . . . . . . . . . . . . . . . . . . . . . . . . 18

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Risk and Resilience. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

viii

Risk and Resilience Misconceptions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

Student-Deficits Versus Student-Assets Orientation. . . . . . . . . . . . . . . . . . . . . . . 22

Generalist Approach. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Bolstering Resilience. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

Caring Relationships. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

High Expectations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Meaningful Opportunities for Participation. . . . . . . . . . . . . . . . . . . . . . . . . . . 26

Internal Student Assets.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

Problem-Solving Skills. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Autonomy.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

Social Competence.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

Sense of Purpose and Bright Future.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

External Protective Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

School as a Protective Factor. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

Emerging Field of Educational Resilience.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

Studies in Educational Resilience. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

Educational Leadership Practices Aimed at Changing School Culture. . . . . . . . . 42

Social Justice Leadership. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

Systems Orientation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

Change Leadership.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

Educational Policy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

School Culture. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

Culture Change. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

No Child Left Behind Impact.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

Testing Orientation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

Assessment Policy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55

Present School Reform Initiatives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

California Healthy Kids Survey. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

Assessing Resilience and Youth Development in Schools. . . . . . . . . . . . . . . 59

Resilience, Youth Development, and Academic Performance. . . . . . . . . . . . 59

School Connectedness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

Middle School Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68

Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

CHAPTER 3—METHODOLOGY. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70

Research Questions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

Instrumentation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71

Population. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

Research Design.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Data Analysis Procedures. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76

Ethical Issues. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77

CHAPTER 4—RESULTS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78

Descriptive Statistics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

Academic Performance Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79

External School Protective Factor Variables. . . . . . . . . . . . . . . . . . . . . . . . . . 79

Internal Student Asset Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80

Demographic Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

School Connectedness Variable.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83

Simple Correlations With Academic Performance Index. . . . . . . . . . . . . . . . . . . 83

External School Protective Factor Variables. . . . . . . . . . . . . . . . . . . . . . . . . . 83

Internal Student Asset Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

School Demographic Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

School Connectedness Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

Correlation Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87

Hierarchical Multiple Regression. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88

External School Protective Factor Variables. . . . . . . . . . . . . . . . . . . . . . . . . . 89

Internal Student Asset Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91

School Demographic Variables. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

School Connectedness Variable.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95

Hierarchical Multiple Regression Summary. . . . . . . . . . . . . . . . . . . . . . . . . . 96

School Connectedness Mediator Model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97

Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98

CHAPTER 5—SUMMARY, CONCLUSIONS, AND RECOMMENDATIONS. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100

Extension of Previous Research.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

Summary of Findings.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

External School Protective Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

Internal Student Assets.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104

School Connectedness. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

Surprising Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

General Discussion of Findings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109

Simple Correlations and Multiple Regressions. . . . . . . . . . . . . . . . . . . . . . . . 109

Further Resilience Investigations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

Implications for Educational Reform and Leadership Practice. . . . . . . . . . . . . . . 113

Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

Sample Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116

Measurement Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

Design Limitations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117

Varied Statistical Analyses. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118

Recommendations for Educational Practice. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 120

School Leader Orientation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

School Culture and School Reform. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123

Student Achievement Variables.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124

School Connectedness Obstacles. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126

Recommendations for Future Research. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

Conclusion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131

xii

APPENDICES

A. Select-Item Survey Questions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147

B. 2004 Regressions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149

C. 2006 Regressions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150

D. 2008 Regressions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

xiii

LIST OF TABLES

PAGE

Table 1. Academic Performance Index Descriptives by Year. . . . . . . . . . . . . . . . . . . 79

Table 2. External School Protective Variables Descriptives by Year. . . . . . . . . . . . . 80

Table 3. Internal Student Asset Variables Descriptives by Year. . . . . . . . . . . . . . . . . 81

Table 4. Demographic Variable Descriptives by Year. . . . . . . . . . . . . . . . . . . . . . . . 82

Table 5. School Connectedness Descriptives by Year.. . . . . . . . . . . . . . . . . . . . . . . . 83

Table 6. Correlations: External School Protective Factors and API by Year. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

Table 7. Correlations: Internal Student Assets and API by Year. . . . . . . . . . . . . . . . . 85

Table 8. Correlations: School Demographics and API by Year. . . . . . . . . . . . . . . . . 86

Table 9. Correlations: School Connectedness and API by Year. . . . . . . . . . . . . . . . . 88

Table 10. Coefficients With School Connectedness by Year. . . . . . . . . . . . . . . . . . . 90

xiv

LIST OF FIGURES

PAGE

Figure 1. The resilience and youth development theoretical framework. . . . . . . . . . . 4

ACKNOWLEDGMENTS

It gives me great joy and pride to be able to recognize those who have been direct

and indirect influences in helping me complete my dissertation journey. The work

represented here is a reflection of compiled life events, forces, and exchanges that have

shaped my growth and development as a learner, teacher, and contributor to the world

around us. Countless people and events, past and present, have contributed to my

success. To those not mentioned directly here, thank you.

First and foremost, a thank you to my wife, Andrea Most, does not begin to

express the gratitude, respect, and love that I have for her. Her support and friendship

over the last 20 years has been unwavering. I thank her and our beautiful children Janey,

Forest, and Bruce. This dissertation is a shared family project in that we have negotiated

our sacrifices of time, sleep, and attention together. Thanks, too, to Elisa Sanchez and

Ben Banatao, whose parenting and direction helped to construct my view of education. In

addition, my extended Sanchez, Washburn, and Wangler families, and my first hero,

Juliana Carandang Sanchez, deserve mention, as their teaching and wisdom have

shadowed me throughout my continued journeys.

Finally, I would like to thank the faculty at San Diego State University, who has

supported me throughout my undergraduate and graduate degrees, culminating with my

deepest respect for my dissertation committee chairperson, Cynthia Uline. She was an

invaluable partner, guide, and resource throughout my entire doctoral program. Thanks,

too, to Peg Basom and Maruta Gardner, who contributed to this project as my dissertation

committee members. Finally, Dan Kitchen and Greg Zarow deserve recognition for their

xvi

support towards a dissertation with a high-degree of quality research. To all, I offer

humbled and sincere thanks.

CHAPTER 1—INTRODUCTION

School accountability measures of the No Child Left Behind (NCLB) Act of 2001

(U.S. Department of Education [USDOE], 2002) aim to improve student performance by

mandating incremental yearly performance goals of schools, so that all students perform

at grade-level standards by 2014. Schools that fail to meet their annual goals face a series

of government sanctions. Consequently, educators are increasingly pressured to provide

academic interventions to address student-learning needs. Federal mandates concentrate

school reform efforts on learning standards and standardized test achievement. School

leaders, however, must also address student social-emotional needs, ensure that all

students are ready to learn, and provide student preventions and interventions that will

assist with keeping students engaged and poised to tackle rigorous curriculum and reach

higher levels of student achievement.

Statement of the Problem

Too often educators narrow their curricular focus to test-driven concerns directing

their attention to specific courses and singular standards in order to meet NCLB

expectations. When this is done at the expense of the affective domain, the students’

readiness to learn and their receptivity to learning may be overlooked.

Current educational practice undermines an asset-based model. An assets model

builds on student strengths and engages learning by connecting students to individualized

interests and talents, whereas an NCLB standardized-test-focus prepares students for a

life of tests rather than the tests of life (Elias, 2001; Lepper, Sethi, Dialdin, & Drake,

1997). Too often, educators’ and policymakers’ focus on standardized tests and a

standard curriculum effectively narrow the curriculum, increase instructional pace, create

less engaging classrooms, and ignore student preferences for authentic, hands-on learning

(Certo, Cauley, & Moxley, 2008). In addition, student remediation efforts focus on

student deficits. Students receive test performance labels. In California, for example,

students performing below grade level are referred to as “far below basic” students. The

consequence to poor test performance is often more instructional minutes in core (tested)

classes for struggling students. Often, low-achieving students are separated from peers

resulting in tracking and a labeling of students despite scientific studies that document the

negative effects of segregation, especially for nondominant linguistic and cultural groups

(Nieto, 1992; Oakes, 1985).

Support classes eliminate elective opportunities from a student’s schedule where

the student may encounter feelings of success, creativity, imagination, and school

connectedness (Benard, 2004). High-stakes testing appears to be particularly detrimental

to resilience and youth development (Kohn, 2000; Meier, 2000; Popham, 2001).

Purpose of the Study

This study plans to explore the relationship of external school protective factors

(caring relationships, high expectation messages, and meaningful participation) and

internal student assets (problem-solving, self-efficacy, empathy, and self-awareness)

with student academic achievement. Study findings may address whether a statistically

significant relationship exists between student perceptions of school protective factors

and internal assets with student performance

Research Questions

This study will utilize both descriptive and inferential statistics. Applied

quantitative methodology will explore statewide survey data of the Resilient Youth

Development Module (RYDM) of the California Healthy Kids Survey (CHKS). The

CHKS is a repeated cross-sectional, self-report survey that the California Department of

Education (CDE) has made available to all of California’s school districts as part of the

CDE’s accountability system. Most California schools administer the CHKS biennially.

This study will utilize extant data from the annual administration of the CHKS, to

evaluate the relationship of school protective factors with its component dimensions of

high expectations, a nurturing adult, and meaningful opportunities to participate, in

relation to California’s student achievement composite score for schools, the Academic

Performance Index (API; CDE, 2009). A Resilience Youth Development Module

(RYDM), a component within the CHKS, measures the extent to which students possess

internal resilience assets such as problem-solving, self-efficacy, empathy, and self-

awareness. In addition, a school connectedness variable will be tested as a mediating

variable to examine its effect on other variables (Figure 1).

The following research questions will guide this study:

1. Is there a significant statistical correlation between school protective factors of

caring relationships, high expectations, and meaningful participation to student

achievement?

2. Is there a predictive relationship between student internal assets of problem-

solving, self-efficacy, empathy, and self-awareness with student achievement?

3. Which protective factors and internal assets exhibit the most powerful

correlation with student achievement?

Figure 1. The resilience and youth development theoretical framework. Adapted from

Resilience & Youth Development (para. 18), by WestEd, 2011. Retrieved from

http://chks.wested.org/using_results/resilience.

Summary of the Literature

Resilience is defined as the dynamic process whereby individuals exhibit positive

behavioral adaptations despite significant adversity or trauma. Challenges may include a

combination of emotional, physical, or social stressors. Such stressors, also known as

risk factors, are thought to endanger a child’s ability to develop in a healthy, well-

adjusted way, preventing them from productively contributing to society (Luthar &

Burak, 2000; Masten, 1994; Rutter, 1989).

Empirical studies of educational resilience cut across race, ethnicity,

socioeconomics, age, international borders, and gender. Studies have established that

caring teachers and schools that provide curriculum and instruction that engage students

in active participation and learning, while maintaining high expectations develop students

who demonstrate resilient characteristics (Rutter, Maughn, Mortimore, & Ouston, 1979;

Solomon, Battistich, Kim, & Watson, 1997; Solomon, Battistich, Watson, Schaps, &

Lewis, 2000; Solomon, Watson, Battistich, Schaps, & Delucchi, 1997). Studies where

students reported caring adults and high expectations in their school reflected high

student motivation and positive attitudes toward school, leading to student engagement.

Studies also concluded that positive student behaviors are related to higher achievement

(Freiberg, Stein, & Huang, 1995; Hawkins, Catalano, Kosterman, Abbot, & Hill, 1999;

Wang, Haertel, & Wahlberg, 1993; Waxman, Huang, & Padron, 1997; Waxman, Huang,

& Wang, 1997). Finally, students who expressed strong social bonds with adults, and

with peers, were less likely to disengage from school and more likely to participate in the

life of the school and to achieve (Resnick et al., 1997; Wehlage, Rutter, Smith, Lesko, &

Fernandez, 1989).

This summary of literature describes the theoretical underpinnings of resilience

across several disciplines. Beginning with the history of the study of risk and resilience

and an acknowledgment of common misconceptions in the field, the review establishes

the context for educational resilience. Using Benard’s (1991) broad theoretical

framework of resilience as a construct, the review will consolidate, and define key terms

of resilience. Further, the exhaustive review suggests three protective factors, including

caring relationships, high expectations, and meaningful ways to participate, which

together promote resilience in students. Benard’s model suggests that as these protective

factors reside within families, schools, and communities, such people and places provide

for developmental needs of safety, love and belonging, respect, power, challenge,

mastery, and meaning. When resilience is fostered and engaged, students’ internal assets,

such as social competence, problem solving, autonomy, and sense of purpose, emerge.

These personal strengths reflect resilience on the part of youth (Benard, 2004).

Specifically, this review will concentrate on the school setting as the arena for bolstering

student educational resilience.

Empirical studies in the growing area of educational resilience research illustrate

the importance of school climate along with student perceptions of their learning

environment and their relationship to student performance. The relatively recent,

expanding body of work in the field of educational resilience recognizes the need for

further study to better define the terms and measurement of resilience in the school

context.

Rutter and colleagues (1979) established the notion that schools play a key role in

youth resilience after conducting an epidemiological study over a 10-year period in 12

Inner London comprehensive schools. Interviewed children, whose parents had been

diagnosed with a mental illness, revealed differences regarding how they recovered from

adverse conditions. Rutter was one of the first to suggest that both individual

characteristics and the children’s environment were important protective factors. He

concluded that students coming from disadvantaged families were more likely to

demonstrate resilient characteristics if they attended schools that provided a source of

external protective factors, such as fostering a sense of achievement, academic pressure

and high expectations, attentive and caring teachers, and good-teacher student

relationships (Waxman, Gray, & Padron, 2003).

As one of the first scientists to define resilience, Werner (1986) studied a cohort

of nearly 700 children from birth through adulthood on the island of Kauai, Hawaii,

beginning in the 1970s. These children were raised in families with adverse living

conditions including poverty, alcoholism, and mental illness. Findings from her seminal

work determined that while two-thirds of the children exhibited destructive behaviors in

their later teen years, one-third of the children did not. The productive and successful

children were referred to as “resilient” children (Werner, 1986; Werner & Smith, 2001).

Other studies have examined high schools that exhibited effectiveness in dealing with

at-risk student graduation rates, attendance, and increased literacy. The mixed

methodology study led researchers to conclude that students who identified themselves in

the mainstream of school culture, having established positive relationships with peers and

adults in the school, were less likely to disengage and drop-out of school (Wehlage et al.,

1989).

Another classic work in the field of resilience, Project Competence, studied the

impact of life stressors on the competency levels of elementary school children in two

urban Minneapolis schools (Garmezy, Masten, & Tellegen, 1984). The study findings led

researchers to question why some children did not succumb to adversity, nor develop

negative adaptations. Fundamental understandings about the differences in life

experiences of children from adverse backgrounds, and a general framework for

conceptualizing the study of resilience, spawned from this body of work (Luthar, 2003;

Waxman et al., 2003).

Further studies used academic grades to determine student resilience. Gonzalez

and Padilla (1997) examined factors that contributed to the academic resilience and

achievement of resilient versus nonresilient Mexican-American high school students from

three high schools in California. The study findings suggested that a school’s purposeful

fostering of the resilience construct, specifically caring relationships and meaningful

opportunities to participate at school, may lead to higher grades and greater academic

achievement among its students.

Nettles, Mucherah, and Jones (2000), along with findings from The Center

for Research on the Education of Student Placed at Risk (CRESPAR), examined the

influence of parent, teacher, and school support on students’ resilience. They found that

caring parents, participation in extracurricular activities, and supportive teachers were

beneficial to student academic achievement.

Waxman and Huang (1996) utilized motivation and classroom learning

environment survey data and found that resilient students reported a significantly higher

social self-concept, achievement motivation, and academic self-concept than nonresilient

students (Waxman & Huang, 1996). This study established the relationship between

student perceptions of school, student motivation, and academic performance.

Despite the research that presents’ resilience as a phenomenon that can be

bolstered by schools, school-based programs, strategies, or policies designed to enhance

resilience are relatively new. Bosworth and Earthman (2002) suggested that school

administrator perceptions of resilience impacted the decision-making of school leaders to

pursue resilience-oriented programs and efforts. School leaders recognized the concept

of resiliency as a relevant organizing point for designing school programs and school

environments. Coincidentally, the principles and theoretical frameworks of educational

resilience overlap with widely accepted frameworks and theories of educational

leadership (Theoharis, 2009; Wagner et al., 2006).

One emerging educational leadership theory that draws parallels to theories of

educational resilience is social justice leadership (SJL). Theoharis (2009) suggested a

construct to improve student achievement: increased access to core learning, improved

core learning, and the creation of a climate of belonging. The SJL construct presents a

powerful means to understanding and creating meaningful, equitable, and just school

reform reflecting principles similar to those advanced in resilience construct.

Additional research in educational reform from Fullan (2000) and Wagner et al.

(2006) describe the dynamic process of school change and school improvement. To

underscore the complexity of school change, the review of literature also highlights

research in educational policy and school culture (Deal & Peterson, 2009; Fowler, 2009).

Presently, the United States’ educational system emphasizes testing and accountability

(Zhao, 2009). Given this educational climate and culture, the review of literature

attempts to illustrate the effects of testing by including research that investigates The

People’s Republic of China, a historically test-oriented educational system. After

providing a historical perspective on testing and accountability, along with present

American school reform initiatives, the literature then points to other means of assessing

school climates and student readiness to learn (Benard, 2004; Deal & Peterson, 2009).

Method

This study utilized both descriptive and inferential statistics. Applied quantitative

methodology explored statewide survey data of the Resilient Youth Development Module

(RYDM) of the California Healthy Kids Survey (CHKS). Most California schools

administer the CHKS to meet the requirements of the federal Safe and Drug Free Schools

Communities Act (SDFSCA). Select-item CHKS survey items were utilized to evaluate

the relationship between the external school protective factors of a caring, nurturing adult;

high expectations; and meaningful participation, along with student internal assets of

problem solving, self-efficacy, empathy, and self-awareness to California’s Academic

Performance Index (API) school-composite score. The study tested the school

connectedness as a mediator variable and controlled for demographic differences.

Statistical analyses replicated over three time points will strengthen the findings of

the relationship between external school protective factors, student internal assets, school

connectedness, and school API. A three-step statistical procedure beginning with simple

correlations, then forced-entry hierarchical multiple regressions models, with and without

the school connectedness variable, will allow for an examination of descriptive statistics

which were supported by appropriate inferential statistical tests, to address each research

question.

Sample

This study used the 2003-2004, 2005-2006, and 2007-2008 RYDM data of the

CHKS for all California schools in Grade 7. In addition, this study used 2004, 2006, and

2008 CDE-calculated API scores for all corresponding California schools. Schools with

complete 2004, 2006, and 2008 CHKS data and API scores participated in the study.

Research Design

The study examined extant data and incorporated a correlational quantitative

design that utilized hierarchical multiple regression analysis to illustrate the relationship

between survey data and achievement scores. Regression analysis controlled for student

socioeconomic status, ethnicity, and other demographic information (Huck, 2008).

Limitations

Analysis of relationships using inferential analysis does not determine truths.

Statistical sampling and statistical analysis illustrate significant relationships between

events and the likelihood of occurring phenomena, but cannot with complete certainty

directly attribute the occurrence of one event to another seemingly related occurrence.

Stated simply, the evidence of a correlation between variables does not prove a causal

relationship. The potential affect of another phenomenon, not explored as part of this

study, may certainly exist (Huck, 2008; Popham, 1993).

In addition, the study is limited by its design in that data are accumulated through

student self-reports. Self-reports were based on perception. One student may have

interpreted a survey question differently from another student and may, therefore, have

responded with a different answer. Variance in responses is expected.

Further, the study is limited to a study of seventh grade student responses in

school years 2003-2004, 2005-2006, and 2007-2008. Mean scores and correlations are

generalizable for the surveyed population with the schools serving as the unit of

measurement. Scores cannot be generalized to individual scores.

Variables of Interest

The independent (predictor) variables of this study are the school external

protective factors of caring, nurturing adults; high expectations; and meaningful

participation, and internal student assets: problem-solving, self-efficacy, empathy, and

self-awareness. In addition, demographic variables, along with school connectedness

were accounted for. The dependent (outcome) variable are school API score. The score

is calculated by the CDE, after accounting for schoolwide test performance of students on

state-mandated examinations.

Data Analysis Procedures

The one dependent variable (API scores) and the multiple independent variables

(RYDM of CHKS) were analyzed through hierarchical multiple regression modeling

using the Statistical Package for the Social Sciences (SPSS) version 17.0.

Quantitative data analysis accounted for effect size and observed power (.80) to

strengthen the evaluation of analysis and to decrease the chances of a Type II error (Huck,

2008).

Research Significance

This study’s findings will provide evidence for the relationship between school

resilience scores and school connectedness variables on student achievement.

Statistically significant relationships between the resilience measures and school

connectedness to student achievement may assist educators with decisions in matters such

as curricular offerings, instructional delivery, and professional development. Recognition

of the impact that external school protective factors have on student achievement may

compel school leaders to rethink how to engage, motivate, and create opportunities for

students to learn.

Academic achievement, when measured only by standardized test performance,

may offer limited insights regarding positive student growth and youth development.

This investigation seeks to utilize student perceptions, collected through a widely

administered survey, in the hope that the research can inform future reform efforts that

emphasize bolstered levels of protective factors within the school community, thereby

positively impacting school cultures and student achievement.

Conclusion

No Child Left Behind mandates that all students reach proficiency in tested

subjects like language arts and mathematics by 2014. Our nation, however, faces

challenges. Educational leaders are searching for keys to improved student learning since

the learning needs of all students are not being met. Whether the constructs of

educational resilience and school connectedness are related to student outcomes remains

unclear.

Definition of Terms

Academic Performance Index (API): the cornerstone of California's Public

Schools Accountability Act of 1999; measures the academic performance and growth of

schools on a variety of tested academic measures through standardized tests (Hanson &

Austin, 2003).

California Healthy Kids Survey (CHKS): California Department of Education

(CDE) approved survey instrument to help schools monitor its goals to maintain a safe

and drug free school to meet the requirements of the federal Safe and Drug Free Schools

Communities Act (SDFSCA; Hanson & Austin, 2003).

Caring, Nurturing Relationships: considered one of three external protective

factors that protects students from risk in Benard’s (1991) conceptual framework of

resilience; the term conveys unconditional loving support.

Change Leadership: a transformational improvement process that requires

schools and districts to sharpen their capacities of reflection and encourages leaders to see

more deeply as to why it has been difficult for organizations and individuals to change.

Effective teaching is described with 3R’s: rigorous, relevant, and based on respectful,

trusting relationships (Wagner et al., 2006).

Climate of Belonging: component of the Social Justice Leadership framework that

creates a warm and welcoming school culture that encourages collaborative communities

within each classroom, incorporating social responsibility into the school curriculum

(Theoharis, 2009).

Developmental ecological models: recognition that youth, families, peers,

neighborhoods, schools, communities, organizations, and larger cultural values are linked

and simultaneously shape child development (Whitlock, 2006).

Educational resilience: a subset within the field of resilience research focused on

the domain and context of schools and schooling (Wang, Haertel, & Wahlberg, 1994).

External Protective Factors, also environmental protective factors: resilience

model supports potentially received by students from adults at school, home, or the

community, such as: caring relationships, high expectation messages, and opportunities to

participate and contribute (Benard, 1991).

Hierarchical Multiple Regression Model: statistical analysis technique that allows

the researcher to determine the order that variables are entered into the regression

equation to control for one or more variables. Comparison of a set of regression models

with slightly different variables allows the researcher to examine the contribution of

independent variables entered into the equation first (Huck, 2008).

High expectation messages: considered one of three external protective factors

that protects students from risk in Benard’s (1991) conceptual framework of resilience;

the term refers to the sense of structure and safety through the application of consistent

student-centered rules, perceived as fair, by children. Messages convey a sense of belief

in the achievement of children (Benard, 1991).

High-stakes testing: in education, refers to implications of standardized test

performance. For a student, pass/fail performance on a test may mean the attainment of a

high school diploma. For a school, schoolwide test score performance consequences may

mean sanctions and penalties.

Internal Student Assets: manifested developmental outcomes when resilience is

engaged, such as: problem-solving skills, autonomy, social competence, and sense of

purpose. The personal strengths contribute to the student’s ability to avoid health-risk

behaviors such as alcohol, tobacco, and other drug abuse; teen pregnancy; and violence

(Benard, 1991).

Meaningful Opportunities to Participate (Meaningful Participation): considered

one of three external protective factors that protects students from risk in Benard’s (1994)

conceptual framework of resilience; the term refers to opportunities that develop

autonomy, self-control, and leadership. Opportunities for contribution allow students to

be active contributors to their classrooms, their schools, their family, and in their school

community (Benard, 2004).

No Child Left Behind (NCLB) Act (2001): educational reform initiative, signed

into law by President George W. Bush, characterized by increased accountability, choice,

and performance mandates (Ravitch, 2010).

Resilience Youth Development Module (RYDM): component of the CHKS survey

instrument with scales intended to measure student perceptions of protective factors

found in school, home, and the community. Also includes measures of internal student

assets, traits thought to reflect resilience (Hanson & Austin, 2003).

School climate: theorized to be composed of five domains: order, safety, and

discipline; academic outcomes; social relationships; school facilities; and school

connectedness. Also commonly referred to a school culture (Zullig, Koopman, Patton, &

Ubbes, 2010).

School connectedness: construct related to school climate and improved school

performance characterized by excited, enthusiastic, and engaged learners; where students

felt valued for their input; and where students had feelings about school. Associated

with, and also referred to as, as school culture, school climate, school attachment, school

membership, school sense of belonging, school bonding, school participation, and student

engagement (Hoy & Hannum, 1997; Osterman, 2000; Witherspoon, Schotland, Way, &

Hughes, 2009; Zullig et al., 2010).

School culture: deeper organizational structure reflected and transmitted through

symbolic language and expressive action (Deal & Peterson, 2009).

School protective factors: resilience model supports received by students from

adults at school, such as caring relationships, high expectation messages, and

opportunities to participate and contribute (Benard, 1991).

Social Capital Theory: describes trust, norms, and exchange as critical reciprocal

elements to the development of communities and to the interconnectedness between

youth and adults (Whitlock, 2006).

Social Justice Leadership (SJL): educational leadership approach intended to

improve student achievement through a three-legged approach to social justice and school

reform, which includes: increased access to core learning, improved core learning, and

the creation of a climate of belonging (Theoharis, 2009).

Student achievement: considered to be a measurable indicator of student success

and student attainment of grade level standards expressed in grades, standardized test

scores, and/or student performance measures.

Support Classes: term used synonymously with remediation classes and

intervention programs to improve student acquisition of grade-level standards and

performance on standardized tests.

Testing orientation: an inclination towards measuring and monitoring the success

of students, schools, districts, and systems primarily through standardized tests (Zhao,

2009).

Youth Development theory: emphasizes the importance of understanding that

contexts have profound impact on how young people thrive and that young people are

dynamic, not passive actors in the world they inhabit (Whitlock, 2006).

CHAPTER 2—REVIEW OF LITERATURE

Introduction

The No Child Left Behind Act of 2001 (NCLB) aims to improve student

performance by designing school accountability measures that monitor continual

academic performance of all students (USDOE, 2002). Schools who fail to meet their

goals face a series of government sanctions. Consequently, educators are increasingly

pressured to provide academic interventions to address student-learning needs. Federal

mandates concentrate school reform efforts on learning standards and standardized test

achievement. School leaders, however, must also address student social-emotional needs,

ensure that all students are ready to learn, and provide student interventions that will

assist with keeping students engaged and poised to tackle rigorous curriculum and reach

higher levels of student achievement.

This review summarizes the multidisciplinary concept and theoretical

underpinnings of resilience research. More specifically, the review explores the utility

of resilience theory in the social context of schools. A growing body of educational

resilience research provides positive evidence of resilience theory applications that

supports all students, especially those who have been traditionally considered at-risk of

low academic achievement. The literature also includes empirical studies of successful

resiliency-based school reform efforts that empower students by providing protective

factors associated with student success and educational resiliency.

Too often, educators narrow their curricular focus to test-driven concerns,

directing their attention to specific courses and singular standards in order to meet NCLB

expectations. When this is done at the expense of the affective domain, the students’

readiness to learn and their receptivity to learning may be overlooked (Zins, Elias,

Greenberg, & Weissberg, 2000). This review explores student intervention and

prevention strategies for students identified to be at-risk of poor academic performance.

Moreover, the review identifies unintended negative consequences of stringent NCLB

standards, intending to examine the educational needs of the whole child, such as school

connectedness; and thus, encouraging a broader perspective, necessary within the current

high stakes policy climate. To illustrate a historical, international perspective, the

review examines outcomes from China’s test-oriented educational system. As a final

consideration, the review challenges educational leaders to reexamine the types of

academic interventions provided to children, and points to the need for further empirical

study especially in the applications of resilience theory to the practice of educational

leadership.

Comparatively, there are few empirical studies of resilience and its relationship to

school leadership. However, several parallels exist between aspects of resilience theory

constructs and the empirical and theoretical research of educational leadership.

Risk and Resilience

Throughout a lifetime, humans confront a variety of life challenges. During

childhood and adolescence, threats to student well-being may be temporal or more

pervasive. Challenges may include a combination of emotional, physical, or social

stressors. Such stressors, also known as risk factors, are thought to endanger a child’s

ability to develop in a healthy, well-adjusted way, preventing them from productively

contributing to society. Risk factors may include sexual abuse, gang affiliation, poverty,

discrimination, dysfunctional family dynamics, drug abuse, and teen pregnancy, to name a

few. Resilience is defined as the dynamic process whereby individuals exhibit positive

behavioral adaptations despite significant adversity or trauma (Luthar & Burak, 2000;

Masten, 1994; Rutter, 1989).

The study of risk and resilience has its roots in psychology. One of the first

scientists to refer to resilience was Werner (1986). Her seminal work beginning in the

1970s on the island of Kauai, Hawaii, studied a cohort of nearly 700 children from birth

through adulthood, raised in families with adverse living conditions including poverty,

alcoholism, and mental illness. While two-thirds of the children exhibited destructive

behaviors, such as substance abuse and teen pregnancy in their later teen years, one-third

of the children did not. The productive and successful children were referred to as

“resilient” children (Werner, 1986; Werner & Smith, 2001).

As risk and resilience research continued, and as various “at-risk” cohorts in their

youth aged into adulthood, researchers found that many of their risk-exposed subjects

had stable jobs and marriages, were satisfied with their spouses and children, and were

responsible citizens in their communities. Longitudinal studies concluded that 70-75%

of those who are challenged by adverse conditions in their youth will be successful by

midlife (Rhodes & Brown, 1991; Rutter, 1987, 2000; Werner, 1986; Werner & Smith,

2001). Resilience research developed a shift in focus from risk factors to “protective”

factors, describing how populations at-risk were able to buffer themselves from failure

and harm.

Werner and Smith (1992) matched their findings with other American and

European investigators who applied a life-span perspective and suggested that protective

factors “make a more profound impact on the life course of children who grow up under

adverse conditions than do specific risk factors or stressful life events. They [also] appear

to transcend ethnic, social class, geographical, and historical boundaries” (p. 202).

Masten (2001) goes on to say, “Resilience does not come from rare special qualities, but

from the everyday magic of ordinary, normative human resources in the minds, brains,

and bodies of children, in their families and relationships, and in their communities”

(p. 9).

The growing body of resilience research has illustrated how severe environmental

risk factors do not predetermine the quality and contributions of one’s life. The concept

and nature of resilience, however, is often misunderstood.

Risk and Resilience Misconceptions

Several misconceptions regarding risk and resilience exist. Some researchers

have viewed resilience as a personality trait that some children have and some do not

(Work, Cowen, Parker, & Wyman, 1990). Others maintain that if most children have the

capacity to overcome hardship, social and political advocacy to promote resilience is

unwarranted. Collected stories about “invincible kids” popularize and distort the notion

of resilience (Brownlee, 1996).

Another related and pervasive misconception of resilience research is that its

findings are applicable to only “at-risk,” or “high-risk,” youth. Therefore, an orientation

towards a focus on risks persists. Risk-focused researchers have identified personal

attributes, family situations, and community features as factors that correlated to drug use,

dropping out of school, or criminal activity (Blum et al., 2000; Brown, 2004; Criss, Petit,

Bates, Dodge, & Lapp, 2002). Although resilience research has noted that risk factors

were only predictive for 20-49% of a high-risk population (Rutter, 1987, 2000; Werner &

Smith, 2001), school and community efforts across the United States concentrate on

addressing risk factors. The United States Government Accountability Office (1997)

estimated that “billions” of dollars were spent on programs for at-risk youth from 1996 to

1998 (Brown, 2004). How did American schools come to concentrate on risk factors?

Student-Deficits Versus Student-Assets Orientation

In 1983, the National Commission on Excellence in Education released its report

titled A Nation at Risk: The Imperative for Educational Reform. The document noted

poor student performance on standardized tests and identified risk factors for student

failure (Brown, 2004). The term “at-risk” entered the vernacular of social scientists and

voluminous evidence correlating risk factors to negative outcomes suggested that some

risky behaviors were thought to predict negative outcomes and life failure (Benard, 2004).

This practice evolved into the notion that our youth are “at-risk” for some type of failure

(Brown, 2004).

Researchers consulting with policymakers created and implemented policies and

programs to assist at-risk and high-risk youth across the nation. State and federal funding

mechanisms compelled school officials to identify the number of risk-factors and at-risk

students present in school communities. The more students and risk factors were

identified, the more much-needed discretionary money schools and school districts

received. This funding dynamic cemented a deficit perception of young people (Brown,

2004).

Two prominent branches of risk and resilience theory have emerged: the

generalist and specifist approaches. The specifist approach is focused on risk-behaviors

and applies a deficit-based, risk-orientation outlook on children (Hawkins, Catalano, &

Miller, 1992).

Hawkins et al. (1992) identified 17 risk factors associated with alcohol, tobacco,

and other substance abuse among adolescents. The researchers concluded that when

more risk factors are present, the greater the likelihood of young people engaging in

alcohol, tobacco, and substance use. Subsequently, identification of risky environments,

or groups of adolescents possessing risk factors, became a focus to prevent life failures,

such as dropping out of school. Risk factors included demographic, family, and

individual personality traits. Exposure to more risk factors meant a greater need for more

protective factors to offset the stressors. As a result, research has focused on eliminating

the factors related to substance use (Farmer et al., 2003; Hawkins et al., 1992).

In recent years, the risk model has come under scrutiny. School and teacher

expectations for an entire group, thought to be at-risk, may be lowered. Standardized test

scores, reported and disaggregated by income level, race, ethnic origin, or language skill

perpetuates the potential that students may be grouped and labeled as disadvantaged

(Calabrese, Hummel, & San Martin, 20070; Catterall, 1998). In addition, since

individuals react differently in different contexts, what may be an obstacle or challenge

for one child may be an opportunity to rise to the challenge for another (Liddle, 1994).

Generalist Approach

Conversely, a generalist approach, a more positive, asset-based orientation,

suggests that resilience is inherent within all human beings and can be deliberately

fostered to create a healthier, well-adjusted, productive person (Luthar, Cicchetti, &

Becker, 2000; Waxman, Brown, & Chang, 2004). Through this generalist approach, the

focus on student resilience shifted from deficit and disadvantage to growth and strength

development. Cefai (2008) described the shift as “[asking] ‘What makes children in

difficulty achieve and be successful?’ rather than ‘What prevents children in difficulty

from succeeding?” (p. 21).

Currently, the most prevalent and widely implemented school intervention

strategies attempt to assist students beyond the inevitable risk factors they will encounter,

thus advancing a deficits-based approach that presupposes all students are at-risk. This

more negative approach undermines efforts to support our students towards wellness and

higher levels of student achievement. These risk-based programs apply social influence

on students. The programs utilize fear arousal, coercion, and rewards to direct students

away from risky behavior. Each of these strategies, however, proves to be ineffective in

the long term (Brown, D’Emidio-Caston, & Pollard, 1997).

The generalist branch of risk and resilience research advocates an asset-based

approach. Among proponents of a more positive approach to resilience, Benard (1991)

asserted “the development of human resiliency is none other than the process of healthy

human development” (p. 18). Adverse conditions did not create a special attribute like

resilience, rather it is a quality developed in favorable and unfavorable conditions that

benefits all humans (Masten, 2001; Masten & Coatsworth, 1998). Benard (1991)

advocated an asset-based approach, in moving from a model of risk to resilience. In

Fostering Resiliency in Kids: Protective Factors in the Family, School, and Community,

she identified the three “environmental protective factors” of caring relationships, high

expectation messages, and opportunities for participation and contribution, as the key

elements in families, schools, and communities to prevent student exposure to risk.

Further, the field of resilience research has expanded the applicability of its

strengths-based, generalist perspective to all children. A universal, generalist approach

suggests that the same factors that benefit children in adversity (caring adults, high

expectation messages, and meaningful opportunities to participate) benefit normally

developing, already motivated children, as well (Solomon et al., 2000; Solomon, Watson,

et al., 1997). With this in mind, it is important to note that throughout this review of

literature, references made to bolster student resiliency, benefit all students, including

those thought to be at risk of lower academic performance.

Bolstering Resilience

The theoretical framework developed and proposed by Benard (2004) after an

exhaustive review of literature suggested three protective factors: caring relationships,

high expectations, and meaningful ways to participate. She applied the protective factors

to three areas: families, schools, and communities. These venues serve as places of

support where youth receive support related to their developmental needs of safety, love

and belonging, respect, power, challenge, mastery, and meaning (Benard, 1991; Benard,

2004).

Caring Relationships

This term conveys unconditional loving support. It is delivered through

kindness, trust, and gestures such as a smile (Higgins, 1994). Caring adults interact with

compassion and a respect of youth, validating the child’s identity. An attentive and

supportive adult models, reinforces, and provides constructive feedback to children to

promote healthy intellectual, psychological, and social growth (Eccles & Gootman,

2002). Schools convey caring relationships when every student has a nurturing

relationship with at least one adult at school. In addition, schools and classrooms feel

like learning communities where classroom practices make use of a number of small-

group processes like collaborative learning, peer helping, and peer support.

High Expectations

This term refers to the sense of structure and safety through the application of

consistent rules, perceived as fair, by children. At schools, high expectation messages

are positive and student-centered. The messages convey a sense of belief in the hopes,

dreams, interests, and achievement of children. When adults deliver high expectation

messages, they allow youth to see themselves as capable, with a sense of purpose, and a

bright future (Benard, 1991). A characteristic of schools that are closing the achievement

gap is a refusal to dumb down or limit opportunities for lower achieving students (Wang

& Reynolds, 1995). Schools exhibiting high expectations provide a rich, rigorous, and

equitable curriculum that accommodates a broad range of students, and where learning

opportunities are structured so that success is possible.

Meaningful Opportunities for Participation

This protective factor is a natural by-product of fostered relationships based on

care and high expectations (Benard, 2004). Opportunities for participation in group-

focused, or cooperative activities, address a child’s psychological need for belonging.

Meaningful participation may also include a student’s opportunity to discuss topics

important to them, such as sexuality and drug use. Adolescents identify a need to voice

their realities in school, family, and in their communities (Brown & D’Emidio-Caston,

1995; Whitlock, 2006). Opportunities for discussion foster critical thinking skills

and empower youth towards sound decision-making. In addition, opportunities for

participation include opportunities to problem-solve and to make school decisions. These

types of opportunities develop autonomy, self-control, and leadership (Benard, 2004).

Finally, opportunities for contribution allow students to be active contributors to their

classrooms, their schools, their family, and in their school community. Students can

reframe their role from being a problem, or receiver of support, to being a provider of

services if given the training and opportunity to serve as a peer helper, for example.

Personal strengths of social competence, problem-solving, a positive sense of self, and a

positive outlook on the future develop from meaningful opportunities to contribute and

participate (Benard, 2004). Students benefit from having voice and choice in their daily

life at school, participating in many relevant experiential learning opportunities, and

pursuing opportunities of community service learning.

Schools provide students with a social setting where they can develop protective,

nurturing supports, especially in the absence of positive family relationships (Rutter et al.,

1979). Schools that provide students with caring relationships, high expectations, and

meaningful ways to participate engage student developmental needs and enhance personal

resilience strengths. These individual traits are also called internal assets or personal

competencies. The internal assets do not cause resilience, but rather, are illustrative of

the positive developmental outcomes of resilience (Benard, 2004).

Internal Student Assets

When students are nurtured in their environments, encouraged and allowed to

develop their basic human needs, these experiences promote individual resilience

strengths, namely: problem-solving skills, autonomy, social competence, and sense of

purpose. They are the manifested developmental outcomes when resilience is engaged.

Development of resilience strengths improves student social interactions, health, and

academic outcomes. Furthermore, the personal strengths contribute to the student’s

ability to avoid health-risk behaviors such as alcohol, tobacco, and other drug abuse; teen

pregnancy; and violence (Benard, 1991).

Problem-Solving Skills

This personal strength in resilience research is often referred to as “good

intellectual functioning” (Masten & Coatsworth, 1998). Other traits are encompassed

within problem-solving skills, including: planning, flexibility, resourcefulness, critical

thinking, and insight. Critical thinking and insight are especially important because they

instill a consciousness that structures of oppression (an abusive parent, insensitive school,

or experiences with racism) can be overcome. These abilities prevent internalized

oppression and a sense of victimhood (Freire, 1993). Insight allows children to realize

that not all fathers are abusive, that a parent’s erratic moods are not normal, and that other

children have different adverse circumstances (Wolin & Wolin, 1993).

Autonomy

This personal strength underlies intrinsic motivation and has profound impacts

on teaching and learning. Autonomy is associated with positive health, a sense of well-

being, a true sense of self and identity, and one’s feeling of power (Deci, 1995).

Autonomy also includes several psychological functions such as: positive identity,

internal locus of control, initiative, self-efficacy, mastery, adaptive distancing, resistance,

self-awareness, and mindfulness.

Positive identity is often used synonymously with positive self-evaluation or

self-esteem. It is consistently used to describe resilient children who have overcome great

odds (Masten & Coatsworth, 1998; Werner & Smith, 1992). Strong positive ethnic

identity is associated with high self-esteem, a strong commitment to do well in school,

a strong sense of purpose in life, and high academic achievement (Eccles & Gootman,

2002). A student’s positive outlook of self fosters a pride in learning and promotes

positive growth and development in school (Benard, 2004). A school, for example, may

promote autonomy when a student’s learning may be personalized to suit her academic

interests and career goals. In addition, a student may decide how she would like to fulfill

a community service project by selecting and supporting her own area of interest.

Autonomy includes a generalized sense of feeling in control or having personal

power, also known as having an internal locus of control (Werner & Smith, 1992).

Resilient students must also recognize what is not in their control, or out of their “sphere

of influence” (Stohlberg & Mahler, 1994). For example, abused children, or students

who are discriminated against at school must not feel as if their mistreatment was their

fault, rather, that it was beyond their sphere of control. Initiative aligns with a locus of

control. It is associated with motivation from within to direct effort and attention towards

a challenging goal (Larson, 2000; Miller, 1990). Larson (2000) sees initiative at the heart

of other strengths such as creativity, leadership, altruism, and civic engagement.

Self-efficacy and mastery is also included within Benard’s review of the construct

of autonomy. Self-efficacy is the belief in one’s power to determine personal outcomes

(Bandura, 1995). Mastery, the feeling of doing something well or feeling competent, is

associated with self-efficacy. Mastery experiences help to develop a sense of efficacy

(Benard, 2004). When people experience feelings of success, they believe that they have

the skills to succeed and will be more ready to bounce back from setbacks or failure

(Bandura, 1995).

Adaptive distancing refers to the power of children to separate themselves from

negative situations or conditions, realizing that they are not at fault for the situation, and

that their life will be different. For example, if a parent is an alcoholic, abusive, or

mentally ill, a child can emotionally detach him or herself from the dysfunction.

Resistance is a form of adaptive distancing (Beardslee, 1997). It is the refusal to accept

negative messages about one’s self, gender, race, or culture. It is another powerful

construct within autonomy.

Self-awareness and mindfulness includes observing one’s thinking, feelings,

moods, and strength with the ability to step back from one’s emotions. Resilience

researchers refer to this attribute as a transformative, reframing power that is the essence

of resilience (Beardslee, 1997; Benard & Marshall, 1997; Wolin & Wolin, 1993).

Social Competence

This personal strength “includes the characteristics, skills, and attitudes essential

to forming relationships and positive attachments to others” (Benard, 2004, p. 14).

Attainment of social competence also includes exhibiting responsiveness, social

communication skill, empathy, caring, compassion, altruism, and forgiveness (Benard,

1991). Goleman (1995) refers to social competence as one of the five components of

emotional intelligence, while Gardner (1993) refers to it as “interpersonal intelligence”

among his original multiple intelligences.

Sense of Purpose and Bright Future

This strength is related to the deep belief that one’s life has meaning. Werner and

Smith (1982, 1992) describe one’s sense of purpose as the most powerful asset to propel

young people toward healthy outcomes despite life’s challenges. Combined with positive

self-identity, a strong, positive future-focus is associated with academic success and

fewer health-risk behaviors (Masten & Coatsworth, 1998; Wyman, Cowen, Work, &

Kerley, 1993).

Goal direction, achievement motivation, and educational aspirations, associated

assets categorized within a sense of purpose, are attributed to student success in school

(Anderman, Austin, & Johnson, 2002). In addition, these assets are attributed to those

who do not abuse alcohol and other drugs, and do not drop out of school despite multiple

risks (Masten, 1994; Watt, David, Ladd, & Shamos, 1995; Werner & Smith, 1992;

Wigfield & Eccles, 2002). Achievement motivation is linked to academic success, such

as, higher rates of high school completion, increased college enrollment, increased math

and reading achievement scores, and higher grades (Scales & Leffert, 1999).

Other sense of purpose attributes includes having a special interest or hobby,

creativity, and imagination (Werner & Smith, 1982, 1992). Engaging in a special interest

to activate one’s creativity and imagination can result in “flow.” Csikszentmihalyi (1990)

describes flow experience as such intense concentration and engagement in a task that

one transcends current challenges, and stresses become distant. Flow theory suggests that

the activity is so gratifying that it creates happiness, life satisfaction, and intrinsic

motivation. Schools, for example, may offer classes through a curriculum of relevant

learning experiences that deeply engage students.

In summary, when resilience is fostered and engaged, personal strengths such as

problem-solving, autonomy, social competence, and sense of purpose emerge. In short,

these personal strengths comprise resilience in youth (Benard, 2004).

External Protective Factors

Benard’s (1991) theoretical framework suggested that when schools, families,

and communities create caring relationships with children, provide opportunities for

participation and contribution, and maintain high expectations, youth flourish into

healthy, happy, and productive persons. Healthy students acquire personal strengths and

exhibit resilience. When students are resilient they are receptive to learning, notice the

care and encouragement modeled by adults, and develop better relationships for

themselves and for those around them. These students will not only be ready for higher

levels of student success, they will also be less likely to engage in health-risk conditions

(Benard, 1991, 2004).

School as a Protective Factor

School environments can bolster student resilience. Schools, especially teachers,

have a major responsibility and lifelong impact on the future of students. Especially

when students do not have a nurturing home life, schools provide caring and nurturing

supports that can change a child’s life of risk to resilience (Masten & Coatsworth, 1998;

Rutter et al., 1979; Werner & Smith, 1982, 1992).

Emerging Field of Educational Resilience

Given that children can manifest resilience, competence, and high levels of

functioning within one domain while not in another, the study of educational resilience,

a subset within the field of resilience research has emerged (Luthar et al., 2000; Wang

et al., 1994). For example, Kaufman, Cook, Arny, Jones, and Pittinsky (1994) studied

students with histories of maltreatment. Almost two-thirds of the students were

academically resilient, while only 21% manifested resilience in the domain of social

competence. Over the last several decades, multidisciplinary empirical resilience

research from the fields of developmental psychopathology, psychology, sociology, and

anthropology has addressed the utility of resilience as a scientific construct. There are

comparatively few, however, conceptual and empirical studies in educational resilience.

Nonetheless, studies of educational resilience have earned recognition as a framework for

examining why some students are successful in school, while other students from the

same socially- and economically-disadvantaged backgrounds are not (Hupfeld, 2007;

Luthar et al., 2000; Waxman et al., 2003).

The following section briefly outlines classic empirical studies of educational

resilience that underscore the school’s impact on youth development. The research begs

why some students are able to overcome adversity, while others do not? Further, a more

complete presentation of recent empirical studies in educational resiliency, focused on the

differences between resilient and nonresilient students and their perceptions of the

classroom and school environments, will follow.

Studies in Educational Resilience

Empirical studies of student resilience cut across race, ethnicity, socioeconomics,

age, international borders, and gender. Studies have established that when caring

teachers and schools provide curriculum and instruction that engage students in active

participation and learning while maintaining high expectations, their students are more

apt to demonstrate resilient characteristics (Padron, Waxman, & Huang, 1999; Rutter

et al., 1979; Solomon, Battistich, et al., 1997; Solomon et al., 2000; Soloman, Watson,

et al., 1997). Where students reported caring adults and high expectations in their school,

they also demonstrated motivation and positive attitudes, leading to high levels of

engagement in learning. Studies also concluded that more positive student behaviors are

related to higher achievement (Freiberg et al., 1995; Hawkins et al., 1999; Wang et al.,

1993; Waxman, Huang, & Padron, 1997; Waxman, Huang, & Wang, 1997). Further,

students who expressed strong social bonds to adults, and with peers, proved less likely

to disengage from school and more likely to participate in the life of the school and to

achieve (Resnick et al., 1997; Wehlage et al., 1989).

Rutter et al. (1979) conducted an epidemiological study over a 10-year period in

12 Inner London comprehensive schools. Through extensive interviews, he studied

children whose parents had been diagnosed with a mental illness and concentrated on

the differences of individuals in recovery to the adverse conditions. They found that the

children escaped relatively unharmed and did not exhibit maladaptive behaviors, nor were

they mentally ill themselves. Rutter et al. were some of the first to suggest that both

individual characteristics and the children’s environment constituted important protective

factors. They further concluded that students coming from disadvantaged families were

more likely to demonstrate resilient characteristics if they attended schools that provided

a source of external protective factors, such as fostering a sense of achievement, academic

pressure and high expectations, attentive and caring teachers, and good teacher-student

relationships (Waxman et al., 2003). Rutter et al. established the notion that schools play

a key role in youth resilience.

Wehlage and colleagues (1989) studied 14 alternative high schools that exhibited

effectiveness in dealing with at-risk student graduation rates, attendance, and increased

literacy. The mixed methodology study led Wehlage and his team to construct a

theoretical framework of school membership (Hagborg, 1994). They found that students

who identified themselves in the mainstream of school culture, and had established

positive relationships with peers and adults in the school, were less likely to disengage

and drop out of school. The study suggested that students were more likely to achieve

and participate at school when positive social bonds with those at school were present.

Further, the study characterized teacher culture at the successful schools by a moral

obligation to serve young people (Cefai, 2008). Together, the studies of Rutter et al.

(1979) and Wehlage et al. (1989) suggest that schools, and the social constructs within

them, profoundly affect the success, or failure, of students.

Another classic work in the field of resilience is the Project Competence study

(Garmezy et al., 1984). The work began by examining the impact of life stressors on

the competency levels of 612 elementary school children in Grades 3-6 in two urban

Minneapolis schools. The longitudinal study did not involve a high-risk sample. Rather,

it was designed to examine competence among a normative school cohort with various

kinds and levels of adversity. In collaboration with researchers, the school superintendent

and principals selected the sample to reflect the diversity of socioeconomic status (SES)

and ethnic minorities within the public school district at the time. Garmezy and

colleagues (1984) directed their attention towards the relationship between competence,

adversity, internal functioning, and a collection of individual and family attributes. In

sum, 205 children and families participated in the follow-up studies at 7, 10, and 20 years

to provide longitudinal data on competence and what later became known as resilience

(Luthar, 2003). During school age years, teacher ratings, peer assessments, and school

record data assessed competence, while stress exposure was measured by a life event

questionnaire. Researchers also interviewed parents about the social structure of the

family and their perspective of their child. Using an exploratory multiple regression

correlation analysis, Garmezy and colleagues discovered that disadvantaged children with

lower IQs and SES, and less positive family qualities, were generally less competent and

more likely to be disruptive in school. However, the researchers found that some of the

disadvantaged children were competent, were doing well, and did not display behavioral

problems. This finding led researchers to question why some children did not succumb

to adversity and did not develop negative adaptations. Key understandings about the

difference in the lives of children from adverse backgrounds, as well as a general

framework for conceptualizing and operationalizing the study of resilience, resulted from

this body of work (Luthar, 2003; Waxman et al., 2003).

It is important to note that these aforementioned classic studies of resilience

illustrate the substantial variation in operationalization and measurement of key

constructs in the field of resilience. For example, when the students of the Project

Competence study (Garmezy et al., 1984) matured beyond school-age years, researchers

developed new criteria to measure whether or not a person was functioning well. As

individuals grow older, new domains of competence become more salient and the

diagnostic criteria to measure one’s success (competence) change. Similarly, identifying

resilience from explicit, or implicit, diagnostic criteria does not describe a person in

totality, nor define their lives at all times. An appropriate indicator of resilience in a

school-age child might be academic success, whereas, an adult-age indicator might be

steady employment. Resilience, therefore, is not a trait of an individual; rather,

individuals manifest resilience in their behavior and life patterns through a dynamic

process (Luthar & Burak, 2000; Luthar et al., 2000). Further, the Garmezy et al. (1984)

study worked with students from uniquely varied, multiple, stressful life conditions.

This, too, illustrates the variety in resilience research and reflects the lack of consensus

about definitions, with variations in operationalization and measurement of key

constructs. The recently expanding field of educational resilience lends greater precision

to the terminology and to the many multidisciplinary spheres of resilience by

concentrating on the particular domain of schools (Luthar & Burak, 2000).

In a study that used academic grades as criteria for resiliency, Gonzalez and

Padilla (1997) examined factors that contributed to the academic resilience and

achievement of 133 resilient and 81 nonresilient Mexican-American high school students

from three high schools in California. High- and low-achieving Mexican American

students were characterized as resilient and nonresilient students, respectively.

Participants completed a questionnaire. Responses formed the study independent

variables while grade point average (GPA) represented the dependent variable. The

selected resilient students reported that their grades so far in high school were “mostly

A’s.” The identified nonresilient students described their grades in high school were

“mostly D’s” or “mostly below D’s.” An ANOVA revealed that resilient students had

significantly higher perceptions of family/peer supports, teacher feedback, positive ties

to school, value placed on school, and peer belonging than did nonresilient students. In

addition, through regression analysis researchers concluded that students’ sense of

belonging to school was the only significant predictor of academic resilience. The

study findings suggest that a school’s purposeful fostering of the resilience construct,

specifically caring relationships and meaningful opportunities to participate at school,

may lead to higher grades and greater academic achievement among its students. A

limitation to this study is the use of grades as a measurement of resilience. Grades

of “mostly A’s” do not equate to resilience, just as poor s do not equate to nonresilience.

Rather, s may more accurately reflect proficiency in an area of study. Further, limiting

the scales of resilience to “mostly A’s” excludes satisfactory s of B and C. In general, the

study’s findings may more accurately reflect the reasons why some students received

better s than others, not resilience.

The Center for Research on the Education of Student Placed at Risk (CRESPAR)

has participated in several studies of educational resilience. An example of a CRESPAR

study involved a longitudinal mixed methodology study of Chicago student transitioning

from the smaller setting of elementary school to high school (Roderick et al., 1997).

Nettles and colleagues (2000) reviewed this and other recent CRESPAR studies that

examined the influence of parent, teacher, and school support on students’ resilience.

They found that caring parents, participation in extracurricular activities, and supportive

teachers were beneficial to student academic achievement. Developing their own

research, Nettles and colleagues studied 75 African American fourth and fifth graders.

The researchers found that students’ perceived exposure to violence had a significant

relationship on mathematics and reading achievement, while teacher support had a

positive impact on mathematics achievement. These findings supported, and were

consistent with, the previous findings of CRESPAR research. The impact of external

protective factors leading to student resilience and increased academic achievement

continued as a valid scientific construct. A criticism of this study, however, centers on

the fact that the relationship between student achievement and teacher support did not

apply to reading achievement as it did for mathematics. This finding indicates that

another factor for improved mathematics performance, such as teacher expertise, may be

more at work than the protective factors of resilience. Otherwise, gains would have been

generalized to reading achievement as well.

In a series of studies conducted by the U.S. Department of Education National

Research Centers; the Center for Education in the Inner Cities (CEIC); and the Center

for Research on Education, Diversity & Excellence (CREDE), researchers examined

differences between resilient and nonresilient elementary and middle school students.

Students were from several culturally and linguistically diverse urban school districts

and lived in low socioeconomic circumstances. In an initial study, Waxman and Huang

(1996) compared the motivation and classroom learning environment of 75 resilient

versus 75 nonresilient sixth-, seventh-, and eighth- students from an inner-city middle

school located in the south central region of the United States. Educationally resilient

students were defined as students who scored at, or above, the 19th-percentile on

standardized achievement mathematics tests over a 2-year period. Nonresilient students

were defined as students who scored at the 10th-percentile, or lower, on standardized

achievement tests over a 2-year period. Three standardized, student self-report

instruments were used to examine students’ perceptions of their classroom learning

environment: the Multidimensional Motivation Instrument (MMI), the Classroom

Environment Scale, and the Instructional Learning Environment Questionnaire.

Multivariate analysis and univariate post hoc tests revealed that resilient students were

found to have significantly higher perceptions of involvement, task orientation, rule

clarity, satisfaction, pacing, and feedback than nonresilient students. Resilient students

also reported a significantly higher social self-concept, achievement motivation, and

academic self-concept than nonresilient students (Waxman & Huang, 1996). This study

fortified an attempt to examine the construct of resilience over time based on consecutive

years’ performance on standardized test achievement data (Iowa Test of Basic Skills),

while correlating performance data to student survey perceptions of their learning

environment and students’ motivation (Waxman & Huang, 1996). Once again, however,

researcher constructs of scales at 19th-percentile and above, versus 10th-percentile or

lower to define resilient and nonresilient students disregards the scores of students in

between. A sampling of students at the extreme ends of an achievement test’s

performance scale does not accurately capture the dynamic process of resilience.

The studies described above illustrate the growing body of research on

educational resilience. Most of the research has been descriptive, comparative, or

correlational. There have been few experimental studies in this area. Padron, Waxman,

Powers, and Brown (2002), however, developed, implemented, and tested a teacher

development program designed to improve resiliency of low-achieving English Language

Learners (ELLs). The Pedagogy for Improving Resiliency Program (PIRP) was

implemented in six fourth- and fifth- classrooms in an urban elementary school serving

predominately Hispanic ELLs from low socioeconomic backgrounds. Yearlong PIRP

training incorporated several components designed to help classroom teachers improve

their instruction and the learning of resilient and nonresilient ELLs (Waxman et al.,

2003).

The findings from the study revealed that the treatment teachers’ classroom

instruction exceeded that of the comparison teachers on some important aspects, such as

providing explanations, encouraging extended student responses, encouraging student

successes, and focusing on the task’s learning processes. Students in the treatment

classes reported a more positive classroom-learning environment than students in the

comparison classes, and they had significantly higher reading achievement gains

than students in the comparison classrooms. Results seemed suppressed, and PIRP

implemented with less fidelity, especially in school districts with a high-stakes testing

focus (Waxman et al., 2003). These experimental findings illustrated the benefits of

purposefully implementing a resiliency-based program to help diverse learners, and also

illustrated that schools with a testing orientation may disregard the benefits of a youth

development approach.

Another quasi-experimental study by McClendon, Nettles, and Wigfield (2000)

examined the effects of Promoting Achievement in School Through Sport (PASS), a

yearlong, elective course in high school, implemented with 900 students from 16 high

schools in the West and Midwest. The PASS classrooms promote protective or resilience

characteristics such as a caring and support, high expectations, and encouragement of

student engagement and involvement. At the end of the school year, students in PASS

were found to have significantly higher s than the comparison group. Classroom

observations revealed that PASS had more indicators of authentic instruction (i.e.,

instructional practices that connect students to meaningful, real-life experiences) than

non-PASS classrooms. Again, PASS illustrates the effect of implementing resiliency-

based programs to improve student s. However, it would be worthwhile to investigate the

program’s effect on standardized achievement scores, as well.

Empirical studies in resilience exemplify the positive effects of s and achievement

scores when students are described as resilient and when resilience-based programs are

implemented in classroom and schools. Few resiliency-based and resiliency-promoting

programs are implemented in our schools (Waxman et al., 2003). However, the

principles and theoretical framework of resilience as it applies to students in school

(providing caring adults, high expectations, and meaningful opportunities to participate)

overlap with widely accepted frameworks for school leadership.

Educational Leadership Practices Aimed at

Changing School Culture

Despite the research that presents resilience as a phenomenon that can be

bolstered by schools, school-based programs, strategies, or policies designed to enhance

resilience are relatively new. Bosworth and Earthman (2002) suggested that school

administrator perceptions of resilience impact the decision-making of school leaders to

pursue resilience-oriented programs and efforts.

Bosworth and Earthman’s (2002) study presented the Henderson and Millstein

(1992) model of resilience to 10 administrators in a large, southwestern United States

city. The 10 participants voluntarily agreed to attend the 90-minute orientation about a

resiliency initiative being implemented by the city. After the orientation, researchers

conducted a semi-structured interview with all participants. The administrators were

encouraged to provide their own definition of resiliency and to provide examples of

resiliency-based approaches they were currently using. Interviewee definitions and

operationalization of resilience were just as varied and vague as the operationalization

and definitions of the resilience research already presented in this review of literature.

Subsequently, five participants agreed to continue with the initiative and with

additional trainings in the Henderson and Millstein (1992) model. A later analysis of

interview transcripts identified common themes across those participants who continued

with the resilience initiative, as compared with those who chose to abstain. Participants

who chose to continue with the resilience training and resilience initiative defined

resilience within an environmental focus with responses such as, “kids need to feel a part

of the community”; “the school should be an environment where kids enjoy coming”; and

“the school should provide meaningful participation.” Conversely, interviewees who

opted-out of the resilience training and resilience initiative described resilience with an

individual focus and as something that resided within the students by suggesting,

“resiliency comes from within” and “resiliency is a spark, a gift” (Bosworth & Earthman,

2002).

The Bosworth and Earthman (2002) study illustrated those individuals who

believed resiliency to be a component of the school environment, tended to continue with

the training; whereas those who believed resilience to be innate opted out. The school

leaders who continued with the training recognized student resiliency to be an

environmental phenomenon that could be promoted within a school’s culture. Thus,

these school leaders recognized the concept of resiliency as a relevant organizing

construct for re-envisioning a school’s culture and designing resilience-focused school

programs and school environments. School leaders with a resilience-focus and

orientation towards providing a positive school environment enhance a positive school

culture. In fact, the principles and theoretical frameworks of educational resilience

overlap with widely accepted frameworks and theories of educational leadership and

organizational change. One particular emerging educational leadership model draws

parallels to theories of educational resilience, namely, social justice leadership (SJL).

Social Justice Leadership

Incorporating, a purposeful and snowball sampling identifying 18 principals, and

with a combination of qualitative methodology along with principles of autoethnography,

Theoharis (2009) utilized principal interviews, a review of documents, site visits, and

focus groups. The 18 principals were identified as having advanced, with success, equity

and justice in their schools. In addition, four criteria were used: they each (a) led a public

school; (b) possessed a belief that promoting social justice was the compelling notion that

brought them to their leadership position; (c) led, kept, and advocated for issues of race,

class, gender, language, disability, sexual orientation, and other marginalizing conditions

at the center of his or her vision; and (d) had evidence to show his or her work had

produced a more just school (Theoharis, 2009). Theoharis operationalized an SJL

construct to improve student achievement and developed a grounded SJL framework with

a three-legged approach to social justice and school reform, which included: increased

access to core learning, improved core learning, and the creation of a climate of

belonging.

The SJL principle of increased access to core learning advances the notion of

student inclusion and opportunity for all. Comparatively, this notion coincides with the

resilience protective factor of high expectations. A school leader, working to increase

access to core learning eliminates pullout and separate programs that work to segregate

students by ability. Rather, an SJL school leader works to increase academic rigor and

access to learning opportunities by increasing learning time. In addition, SJL principals

increase accountability for the achievement of all students (Theoharis, 2009).

Similarly, the SJL principle of improving core learning speaks to the resilience

protective factor of high expectations. Undergirding the SJL principle of improving core

learning is the notion of equity and the belief that all students can achieve. Social justice

leadership describes improvements to core learning in ways that address teaching and the

curriculum. Social justice leadership leaders facilitate teacher development and focus on

equity by addressing issues of race and providing ongoing staff development focused on

building equity. Schools led by SJL leaders adopt common research-based curricular

approaches that empower staff. Theoharis (2009) further described that when SJL leaders

operate through an equity lens, a climate of belonging develops throughout the school

campus. As SJL leaders and teachers strive for equity, students feel that teachers care and

accordingly have high expectations for student success. Comparatively, Benard (2004)

described the resilience protective factor of opportunities for participation and

contribution to be an outgrowth of a school’s caring relationships with high expectations.

Finally, the third component of the SJL approach addresses the creation of a

climate of belonging and is predicated upon connecting and respecting students. Here,

Theoharis’ (2009) model echoes the resilience protective factor of providing meaningful

opportunities for participation and contribution. Social justice leadership leaders create a

warm and welcoming school culture encouraging collaborative communities within each

classroom, in addition to incorporating social responsibility into the school curriculum.

Further, SJL leaders reach out to marginalized families and see community members as

partners to improving student achievement and performance.

To summarize, the SJL construct presents a powerful means to understanding and

creating meaningful, equitable, and just school reform. The model reflects principles

similar to those included within the resilience construct. Social justice leadership

principals keep issues of race, class, gender, disability, sexual orientation, and other

historically marginalizing factors at the center of their practice and vision. Their

leadership necessitates inclusive school practices for students with disabilities, ELLs,

and other students traditionally segregated in schools.

Socially-just leaders eliminate pull-out and self-contained programs for diverse

learners and create inclusive and integrated services whereby children are taught in

heterogeneous groups and receive services from collaborative teams of professionals

within the general education classroom. Thus, SJL challenges the inequity of segregation

and tracking (Theoharis, 2009).

Benard’s (2004) resilience construct depicts how within positive school

environments, students may acquire feelings of caring adult relationships, high

expectations, and meaningful participation. The presence of these environmental assets

can be promoted by school leaders who are aware of how the school’s organizational

constructs and systems affect school culture. Social justice leadership decisions such as

implementing inclusive classrooms, desegregating classes, and eliminating student labels

promote a climate of belonging and build a positive school culture. Successful school

leaders with a holistic, systems-approach to school organization, like SJL, make

leadership decisions that positively impact school culture and build capacity for

successful school change, similar to how a resilience-focused leader acknowledges the

school culture.

Systems Orientation

Fullan (2000) suggested achievement of successful change in elementary school

takes about 3 years. Meanwhile, successful change in secondary schools may take up to

6 years. He further recognized that successful change is happening in only a relatively

small number of schools. Fullan asserted that initial, successful change may be short-

lived. In order to sustain school improvement and successful change, school leaders

must account for the school’s surrounding infrastructure as elements that are critical to

successful change efforts. He provided a metaphor to the bottom-up/top-down

combinations of school organizational change by describing “the three stories of reform.”

The “inside story” describes what is known about how schools change as is applies to

internal dynamics: professional learning communities, a focus on student work through

investigations of assessments, changed instructional practices to yield better learning, and

change to school organization and culture. The “inside-out story” can be described as a

school embracing external forces and pressures to effect changes for the better. The

external forces on schools include: parents and community, technology, corporate

connections, government policy, and the widening teaching profession. The inside-out

belief argues that schools cannot wholly affect school change without embracing these

outside forces. By embracing these threats, a school organization mobilizes resources and

makes its mission coherent. The “outside-in story” describes the school system, or

district, role. The central office role supports the schools by providing autonomy and

decentralizing operations, while at the same time provides encouragement, expertise, and

accountability measures (policy, training, professional development, on-going support) to

allow the schools to develop their local capacity which stimulates innovation. Finally,

Fullan suggested school leaders must recognize the delicate balance between the “three

stories” in order to create cohesion and a unified sense of purpose towards successful

change. Fullan describes a broad, positive school culture, systems-approach to successful

change; much like a resilience-focused leader and socially just leader would approach

school improvement.

Change Leadership

Wagner et al. (2006) presented a similar approach to organizational change. He

described leaders simultaneously sharpening their outward and inward attentions. After

over 5 years of collaboration and study with school and district leaders from all over the

United States, representing urban and rural; large and small school districts; and fiscally

sound and impoverished school districts, Wagner et al. presented a Change Leadership

framework to re-invent and reform American educational practice. They described a

transformational improvement process that requires schools and districts to sharpen their

capacities of reflection and encourages leaders to see more deeply as to why it has been

difficult for organizations and individuals to change. As leaders reflect and inquire more

deeply, they identify the actions necessary to change organizations and individuals.

A Change Leadership approach challenges leaders to create a system for

continuous improvement of instruction and supervision rooted in an organization’s

common vision of effective teaching that is rigorous, relevant, and based on respectful,

trusting relationships. Conceptually, these Change Leadership tenets, also known as the

3 R’s, align to Benard’s (2004) resilience construct that suggests school environments can

promote student resilience by providing students with caring nurturing adults (respectful,

trusting relationships), high expectations (rigor), and meaningful opportunities to

participate (relevance). Building positive school cultures as an avenue to school

improvement cannot succeed without the support of school policies aligned to the

educational mission (Cohen, McCabe, Michelli, & Pickeral, 2009).

Educational Policy

Educational policy can constrain schools and district leader decision-making.

Fowler’s (2009) research on policy implementation considerations cites, “In his

influential book The Culture of the School and the Problem of Change, first published in

1971 and reissued in 1996, Seymour Sarason argued that most education reforms fail

because reformers do not take school culture into account” (p. 272).

Fowler’s (2009) discussion regarding the difficulty of policy implementation

bluntly explains why most reform efforts fail. In this era of accountability and NCLB

mandates, school boards and districts are attempting to discover new ways and best

practices in order to fulfill federal mandates. With this pressure, schools are more apt to

adopt programs designed to meet mandated performance measures in tested subjects,

without weighing the potential effects on school culture (Cohen et al., 2009; Osterman,

2000).

Fowler’s (2009) research underscores the notion that school leaders are key

creators of a school and district’s culture. The establishment of a positive and innovative

organizational culture that is receptive to change and long-term implementation, however,

is often overlooked as an aspect of school reform (Deal & Peterson, 2009).

Proactive leaders, wanting to sustain policy development and implementation, are

wise to consider: Is this the culture we want in our schools? School leaders concentrated

on student achievement through improved standardized test scores, base their

organizational decisions on external forces and the outside mandates of NCLB. Too little

attention has been paid to how schools can be shaped from within (Deal & Peterson,

2009). Similarly, when school leaders limit their attention to student outcomes and high-

stakes testing, they neglect the whole child and the student’s developmental assets which

optimize their readiness to learn. The most successful school leaders consider the school

environment, a positive school culture, and a student’s developmental needs to maximize

student success—the elements of the resilience construct (Benard, 2004; Fullan, 2000;

Wagner et al., 2006).

School Culture

While reformers press for new structures and more rational assessments, it is

important to remember that deep changes cannot succeed without cultural support from

within. Deal and Peterson (2009) describe positive school cultures as places where

shared sets of values support professional development, where a shared sense of

responsibility of student learning pervades, and where a caring atmosphere radiates.

Positive school cultures believe that all students can learn. In addition, positive school

cultures create policies and procedures and adopt policies that support their belief in the

ability of every student (Deal & Peterson, 2009, Fullan, 2001).

In contrast, Deal and Peterson (2009) describe a toxic school culture as one where

teachers are at odds in their belief of every student’s ability to succeed and where a

negative attitude prevails. Educators in toxic cultures see student success based upon the

extent to which students are attentive, concerned, and willing to comply with school

procedures. The policies and procedures of a toxic school support the belief in the

impossibility of school wide student achievement (Deal & Peterson, 2009). For example,

a school master schedule with more remediation classes than college preparatory classes

may signal an institutional belief of low expectations and an obstacle to rigorous

coursework.

Culture Change

Educators focused on the prevailing beliefs and assumptions held by members

of the school organization can better affect cultural change and create positive school

cultures. A distinction must be drawn between technical change and cultural change.

Technical changes refer to changes in school structures, policies, or teaching tools.

Technical changes support professionals and help them do their jobs more effectively.

On the other hand, cultural change is more difficult to accomplish. For example, a school

may implement and adopt a new bell schedule and configure block schedules so that

students can learn. Although the technical change may be necessary, the technical change

produces few positive results when used by people who do not believe in the intended

outcome of the change. Cultural change must precede technical change. Leaders must

address existing assumptions, beliefs, expectations, and habits of the organization if

cultural change is to support the sustainability and success of technical change efforts

(Deal & Peterson, 2009; Muhammad, 2009).

No Child Left Behind Impact

No Child Left Behind reformed education in the areas of standards,

accountability, and choice. A standards-focus ensured a viable curriculum for all

students. In addition, accountability measures through testing and the disaggregation of

student data, focused educators on student subgroup performances and issues of equity in

ways that American educational reform had not done in the past. Public reports put a

spotlight on school performance. Parents were given the choice to move their child from

a low-performing school to a higher performing school, or parents could enroll their

student in a charter school—another outcome from NCLB choice initiatives. Despite

NCLB sanctions and penalties, data show that the achievement gap between African-

American and Latino students compared to White and Asian American students still

exist. Further, only one out of every five charter schools has been shown to outperform

public schools (Ravitch, 2010).

Among the unintended outcomes of federal accountability sanctions is the

narrowing of the taught curriculum and the diversion of students into intervention

programs due to low-test scores. These programs, in essence, separate our lowest

achieving students. The apparent achievement gap between our African-American

and Latino student compared to White and Asian American students, then, perpetuates

tracking and segregation especially since our lowest achieving students become the focus

of pullout and remediation programs (Deal & Peterson, 2009; Muhammad, 2009;

Theoharis, 2009; Zhao, 2009). Leaders with a resilience-focus on students, acknowledge

youths’ need to feel a sense of belonging and care from adults. School leaders too

focused on increasing test results through a separate remediation course limit a student’s

sense of belonging with other students at school and inhibit feelings of success.

Testing Orientation

Current American educational practice emphasizes testing and accountability.

Lessons can be learned, however, from other nations with historic test-oriented

educational systems. The People’s Republic of China is one such educational system that

has historically stressed testing. Early evidence of an Imperial Exam, or Civil Exam,

called keju dates back to 605 AD and the Sui dynasty (AD 581-618). In use for more

than 1,300 years to select government officials throughout the whole nation, it was the

Emperor’s tool to identify and recruit talented individuals to join the ruling class.

Although keju was not part of an educational system, rather a political system, it

determined education in China for centuries because of its high-stakes implications and

power to move an individual, and his family, from one class of society to the next (Zhao,

2009).

Presently, China’s National College Entrance Exam (gaokao) has become as

powerful as the keju. The gaokao acts as a gatekeeper for college and university

admissions and screens Chinese citizens for the opportunity of social and geographic

mobility. Illustrating its continued emphasis towards a test-oriented social and

educational system, a common Chinese expression states, “One exam determines your

whole life” (Zhao, 2009, p. 80).

Keju’s concentration on memorization of the classics came at the sacrifice of

studies in science and technology. By 1905, the emperor issued an order to stop all forms

of the keju exams (Zhao, 2009).

Similarly in 1997, the Chinese Ministry of Education—then the Chinese National

Education Commission—issued a policy against a test-oriented education claiming it

ignored real needs of the student and society; neglected the majority; neglected moral,

physical, aesthetic education; ignored psychological and emotional development; and

relied on rote memorization and mechanical drills.

Contrastingly in the United States, NCLB (USDOE, 2002) mandated an extensive

accountability system involving the state and local education agency. No Child Left

Behind promoted a test-oriented educational system. Specific responsibilities are

assigned to the various agencies, including the local education agency: the school district.

Punitive consequences are explicitly spelled out if agencies fail to fulfill their

responsibilities. States and school districts have developed elaborate systems to collect,

analyze, and report data that are then published in the media, and in other mandated

public documents, such as a school accountability report card (SARC). As NCLB

emerged, parents and community members absorbed the data regarding a waning

American education system in the face of global competition (Zhao, 2009).

Peters and Oliver (2009) asserted that countries implementing high stakes

assessment policies communicate that a productive citizenry is achieved through wide

scale proficiency in reading and mathematics. Moreover, Peters and Oliver (2009) stated

that high stakes testing (a) assumes all students must meet the same standards, (b) fails to

recognize individual differences, talents, and achievements, (c) promotes a culture that

blames, stigmatizes, and excludes students and their teachers, and (d) establishes

mechanisms that all but guarantee segregation, retention, or dropping out of school. The

exclusion and segregation of students is apparent under a system of high-stakes testing.

High stakes assessment in market-driven economies has increased exclusionary practices

(Peters & Oliver, 2009).

Assessment Policy

Further, significant evidence from research studies points to a narrowing of the

curriculum and teachers who are abandoning effective instructional practices to teach to

the test (Peters & Oliver, 2009). Through extensive data collection during classroom

observations, Cuban (2007) found that teacher decisions about textbooks, worksheets,

projects, and many other activities accommodate state tests and accountability

regulations.

A narrowed curriculum eliminates the development of talent in a country.

Richard Florida “documented the increasing importance of creativity and talent for

economic growth” (as cited in Zhao, 2009, p. 51). According to Florida (2005) economic

productivity requires a multitude of talents. Talent diversity breeds innovation and

encourages innovators (Zhao, 2009).

As China takes steps to move beyond a test-oriented educational system,

attempting to avoid negative impacts to creativity and innovation, the United States is

implementing more national control within its educational system aiming to improve

the United States’ world-rank in education (Zhao, 2009). Reform policies of China’s

Ministry of Education reflect a purposeful attempt to provide more local control and

autonomy to schools within each province.

Similarly, the United States has engaged in a series of educational reforms.

However, a perusal of A Blueprint for Reform: The Reauthorization of the Elementary

and Secondary Education Act (USDOE, 2010) does not provide much hope that the

American education system will focus on talent and creativity. In fact, A Blueprint for

Reform sounds eerily similar to No Child Left Behind.

At first glance, A Blueprint for Reform seems to focus on an accountability system

based on rewards. However, the Blueprint construct requires State accountability systems

to recognize progress and growth, and reward success, rather than only identifying failure

(USDOE, 2010). Moreover, districts and states must provide their schools, principals,

and teachers the support they need to succeed.

This sounds like a significant change from the punitive pressure cooker educators

experienced under NCLB (USDOE, 2002). However, close scrutiny reveals that while

rewards for successful schools, districts, and states certainly take a prominent position,

the Blueprint offers an extremely prescriptive solution; some call it a punishment, for

challenge schools that do not succeed immediately.

The first category of challenge schools will be the lowest-performing 5% of

schools in each state, based on student academic achievement, student growth, and

graduation rates. Deriding any flexibility or innovation for improvement, states and

districts will be required to implement one of four school turnaround models (USDOE,

2010).

Schools that are not closing significant, persistent achievement gaps will

constitute another category of challenge schools. In those schools, districts will be

required to implement data-driven interventions to support those students who are

farthest behind and close the achievement gap. For all challenge schools, districts may

implement strategies such as expanded learning time, supplemental educational services,

public school choice, or other strategies to help students succeed (USDOE, 2010). These

phrases appear almost verbatim from NCLB (USDOE, 2002).

Although A Blueprint for Reform (USDOE, 2010) appears heavily laden with

references to standardized tests, accountability, and prescriptive models for low-

performing schools, two areas of the Blueprint provide hope for the American education

system. One section of the Blueprint addresses the successful, safe, and healthy student.

Within that section, the Blueprint does mention the need to provide children with cultural

enrichment. Another section describes a complete education, including: literacy, science,

technology, engineering, mathematics, history, civics, foreign languages, the arts,

financial literacy, environmental education, and other subjects (USDOE, 2010). A

broadening of the taught curriculum allows for more opportunities where students may

encounter feelings of success, connect to the school, and bolster their educational

resilience.

Present School Reform Initiatives

Current educational practice undermines an asset-based model and presupposes

change through mandates with no consideration of the negative impacts of a test-oriented

system. Educators and policymakers narrowly focused only on standardized tests and a

standard curriculum narrow the curriculum, increase instructional pace, create less

engaging classrooms, and ignore student preferences for authentic, hands-on learning

(Certo et al., 2008; Peters & Oliver, 2009; Zhao, 2009). Most student remediation efforts

focus on student deficits, increase instructional minutes in core (tested) classes, and

thereby separate low-achieving students from peers, resulting in tracking, labeling, and

segregating of students despite scientific studies that document the negative effects,

especially for nondominant linguistic and cultural groups (Nieto, 1992; Oakes, 1985).

A popular reform strategy to raise achievement scores involves support classes that

eliminate elective opportunities from a student’s schedule where the student may

encounter feelings of success, creativity, imagination, and school connectedness (Benard,

2004). High-stakes testing appears to be particularly detrimental to resilience and youth

development (Kohn, 2000; Meier, 2000; Popham, 2001). A limited focus on the results

of high-stakes testing might ignore another valuable instrument to school reform, the

California Healthy Kids Survey.

California Healthy Kids Survey

Schools administered the first California Healthy Kids Survey (CHKS) in 1998.

The CHKS is a repeated cross-sectional, self-report survey that the California Department

of Education (CDE) has made available to all of California’s school districts as part of the

CDE’s accountability system, with the recommendation that it be administered biennially.

An advisory committee of researchers, teachers, school prevention and health program

practitioners, and public agency representatives developed the instrument (Hanson &

Austin, 2003).

Most California schools administer the CHKS to meet the requirements of the

federal Safe and Drug Free Schools Communities Act (SDFSCA). The CDE identified

performance indicators that schools must monitor in meeting the SDFSCA’s goals, as

required by NCLB. The CHKS instrument helps schools monitor its goals to keep a safe

and drug free school. In addition, the Resilience Youth Development Module (RYDM)

of the CHKS surveys student perceptions regarding levels of external supports, such as

caring relationships, high expectations, and opportunities for meaningful participation,

within a school, home, and community—also known as protective factors within the

resiliency model. Further, the RYDM attempts to assess internal student assets

of problem-solving, self-efficacy, empathy, self-awareness, cooperation, and

goals/aspirations. The survey provides a common statewide set of comprehensive health

risk and resilience data to guide local program decision-making. The large CHKS

database has the potential to provide critical data needed to examine student learning

outcomes (Hanson, Austin, & Lee-Bayha, 2004).

Assessing Resilience and Youth Development in Schools

Administered to students anonymously, the CHKS and the RYDM provide

school-level data. However, the developing literature surrounding the RYDM and its

normative properties, as they relate to ethnicity, gender, s, and level of family assets,

provide researchers the potential of administering the RYDM’s validated and reliable

subscales to individuals. Research has provided evidence that schoolwide RYDM

subscale responses are correlated to schoolwide student achievement. The potential of

applying the RYDM subscales to individuals in order to assess their potential for

academic success continues to develop (Furlong, Ritchey, & O’Brennan, 2009; Hanson &

Austin, 2003; Hanson & Kim, 2007; Sharkey, You, & Schnoebelen, 2008).

Resilience, Youth Development, and Academic Performance

Concurrent (cross-section) relationship and longitudinal empirical analyses

reveal a relationship between schoolwide RYDM subscales and schoolwide academic

achievement. Researchers collected data and examined the relationship between CHKS

data from 1998-2002 with API research files (calculated from SAT -9, national percentile

rank scores) from 1999-2001. Data were examined two ways: through a cross-sectional

(single time point) analysis and a longitudinal analysis. Stepwise regression models

controlled for demographic differences, such as, socioeconomic status, racial/ethnic

composition, parent education level, free/reduced lunch participation rates, percentage of

English learners, and low or high performing school baseline performance (Hanson &

Austin, 2003; Hanson et al., 2004).

Participants included a total of 800,000 students from s 7, 9, and 11. At the time

of the analyses, schools administered the CHKS voluntarily and the RYDM portion

functioned as an additional voluntary supplemental module. At this time, 1,700 schools

administered the CHKS core module compared to 600 schools who administered the

RYDM module. For the purposes of this discussion, the relationship between the RYDM

module and the CHKS school connectedness scale to student achievement is most

pertinent.

Cross-sectional analyses revealed that API scores were related to school protective

factors. Low-performing schools generally have more students exposed to health risks

and fewer school supports than other schools. In addition, longitudinal analyses revealed

that test score gains were larger in schools with high levels of caring relationships at

school, high expectations at school, and participation in meaningful activities in the

community (Hanson & Austin, 2003; Hanson et al., 2004).

Limitations to the Hanson et al. (2004) study include CHKS data collected prior to

the state mandate in 2004 requiring biennial administration of the CHKS core and

RYDM. Since requiring the RYDM portion of the CHKS, three times as much RYDM

data have been collected per year (Furlong et al., 2009). Other limitations or changes

since the Hanson et al. (2004) study relate to the construct of API scores. Whereas, the

SAT-9 national percentile rank constituted achievement measures at the time of the study,

presently California’s assessment system includes the scaled scores of the criterion-

referenced California Standards test (CST’s). Finally, recent research of RYDM’s

psychometric properties highlights another limitation to Hanson et al.’s study related

to reliability and validity of certain RYDM subscales. Statistical analyses revealed

subscales intending to be conceptually different aspects of resilience measuring to be the

same when verified by a secondary analysis. Therefore, some survey items have been

deemed to be invalid.

To summarize, future research investigating the relationship between RYDM

and API scores should be updated with a larger longitudinal investigation and sampling

of schools that are now mandated to administer the RYDM. In addition, future

investigations should adjust the dependent variables to criterion-referenced scaled scores

versus the norm-referenced national percentile ranks. Finally, future investigations

should utilize psychometrically valid and reliable RYDM subscales.

School Connectedness

Educational research has come to recognize school connectedness as a construct

related to improved school performance. School connectedness has been associated with,

and also referred to as, as school culture, school climate, school attachment, school

membership, school sense of belonging, school bonding, school participation, and student

engagement (Hoy & Hannum, 1997; Osterman, 2000; Witherspoon et al., 2009; Zullig

et al., 2010). Much like the field of educational resilience, despite a growing body of

research to support the role of school connectedness and its relationship to school

improvement efforts, the field lacks a universally-accepted definition of school

connectedness, a commonly-accepted instrument to measure it, and widely-employed

strategies to foster it (Cohen et al., 2009; Hoy & Hannum, 1997). Perhaps, because

studies of school connectedness and school climate are derived from several disciplines,

such as: education, psychology, medicine, anthropology, and sociology; terms overlap,

used inconsistently, and fail to define a clear empirical base of school climate and

connectedness literature (Blum, 2005; Whitlock, 2006).

Studies have attempted to define school connectedness and to construct a reliable

and valid instrument to measure school connectedness. The School Connectedness scale

of the CHKS, for example, has been utilized in hundreds of studies (Whitlock, 2006).

There are several other surveys, too, that seem to measure related constructs, but each

attempts to measure slightly different domains, asks different questions, and thusly,

utilizes different terminology. In addition, several school connectedness measures are

administered without psychometric reliability and consistency studies to validate the

instrument. The absence of a widely accepted, psychometrically-sound instrument has

characterized school climate and school connectedness education research (Zullig et al.,

2010).

Acknowledging that a growing body of research shows school connectedness to

be a powerful predictor of adolescent health and development outcomes, Whitlock (2006)

sought to advance a theoretically grounded definition of school connectedness through a

mixed methods approach that examined the relationship of school connectedness to four

developmental supports: meaningful roles at school, safety, creative engagement,

and academic engagement. The four developmental supports were drawn from

developmental ecological models, social capital theory, and youth development theories.

The theoretical overlap between the theories provided the contextual correlates for the

developmental supports (Whitlock, 2006).

Further, Whitlock (2006) reviewed five surveys, including the CHKS school

connectedness scale, to construct a survey measure that was inclusive of questions meant

to measure: meaningful roles, safety, academic engagement, and creative engagement.

The study utilized the resulting 110-item survey. The survey, however, was not normed

on a larger sample for reliability or validity. The survey findings were drawn from a

sampling of 305 students in the northeastern United States, 83% of whom were European

American. In addition, 3.1% of survey respondents were identified as low socioeconomic

status. Further studies of school connectedness would benefit from the use of a validated

and reliable tool administered to a more diverse population. In order to address reliability

and validity concerns, Whitlock (2006) subsequently assembled focus groups with 108

students, stratified by 8th, 10th, and 12th grades to triangulate survey data and offer

concurrent validity with the quantitative findings of the surveys (Whitlock, 2006).

Whitlock’s (2006) study findings demonstrated that the “meaningful roles at

school” variable most strongly correlated to school connectedness (á = .732, p < .01). In

addition, focus group responses corroborated that “meaningful input into school policies

and practices in and outside of the classroom,” along with “engaging and relevant

classroom class material,” most strongly affected perceived levels of school

connectedness. Considered together, the study findings help to broaden the

conceptualized definition of school connectedness as a mental state of belonging where

youth perceive that they and other youth are cared for, trusted, and respected by adults

whom they believe hold the power to make institutional and policy decisions. In

addition, connectedness is conceptualized as something received and reciprocated. It is

also important to note that study findings suggested that students of all ages struggled

with the relevance of school approaches and curricula in relationship to their interests and

futures (Whitlock, 2006).

Zullig and colleagues (2010) combined findings from several exhaustive

literature reviews of school climate dating back 100 years to reveal at least five important

school climate domains to offer a definition of school climate: order, safety, and

discipline; academic outcomes; social relationships; school facilities; and school

connectedness. Here, school connectedness is a domain within the larger construct of

school climate. The school connectedness domain was characterized by excited,

enthusiastic, and engaged learners; where students felt valued for their input; and where

students had feelings about school (Zullig et al., 2010).

Further, Zullig and colleagues (2010) examined five widely, historically cited

school climate measurement tools. The instruments were matched to the five domains

identified in the literature. The purpose of matching the instruments to the domains

apparent in the literature was to establish validity and reliability measures of the

constructs, and to combine matching survey items to refine a self-report survey from

existing school climate measures with psychometrically tested properties to create a new

survey. A series of exploratory and confirmatory factor analyses tests administered with

several test groups produced an eight-factor model that subdivided social relationships

into three distinct areas: social environment, positive student-teacher relationships, and

perceived exclusion/privilege. The researchers concluded that further development of

their scales would be needed before use as a clinical tool. Although the domains

positively correlated, some of the correlations were negative and very weak. For

example, the range of alpha levels for the perceived exclusion/privilege variable,

appeared to be weak (á = .04, -.10, .04, .04, .02, and .05). In addition, tests with a more

diverse population would further validate the instrument since the tests were conducted

with an 84% sampling of White/non-Hispanic students (Zullig et al., 2010).

Unfortunately, a lack of consistency in terminology may be contributing to a

continued lack of focus toward school connectedness (Whitlock, 2006). In addition,

despite studies suggesting that most local school leaders believe school culture

contributes to increased student performance, state and federal policies fixed on tested

academic performance outcomes increase demands on local leaders to meet student

proficiency mandates, further hampering developmental efforts to establish local policy

that measures and promotes positive school climate and connectedness (Cohen et al.,

2009; Osterman, 2000).

School climate policy. Cohen and colleagues (2009) conducted an investigation

of the relationship between school climate-related findings and educational policy, school

improvement, and teacher education. A historical analysis, a review of literature, a

national State Department of Education scan, and a national survey of school leaders

revealed that despite a growing body of empirical research that indicates the predictive

relationship between positive school climate and academic achievement, there is a gap

between research findings and educational policy and practice (Cohen, 2006; Cohen

et al., 2009).

A review of literature suggests four major areas shape school climate: safety,

teaching and learning, relationships, and the external environment. These four major

areas are composed of several subcomponents that overlap with resilience and social

justice leadership constructs. Namely, the teaching and learning domain is characterized

by elements such as quality of instruction that includes high expectations, real-life

learning, and opportunities to participate. Further, the relationships domain is

characterized by respect for diversity, collaborative learning, and connectedness (Cohen

et al., 2009).

A scan of state departments of education revealed that despite NCLB efforts to

improve schools, and despite the compelling research that suggests school climate is

directly linked to student achievement, many states have left the notion of school climate

out of their general accountability systems. Study findings indicate that 22 states have

integrated school climate into their school improvement and accountability systems;

another 6 did so, as well, but only partially. The other 22 states considered school climate

as a health, special education, or school safety issue. The latter evidence suggests that

despite the research citing school climate to be an integral part of achievement, 22 states

do not relate whole school improvement and academic achievement to school climate. In

fact, only one state department of education, Rhode Island, has formally endorsed or

mandated the use of a research-proven climate assessment; the others are relying on

scientifically unsound assessment tools (Cohen et al., 2009).

Cohen and his team (2009) suggest that the startling gap between the evidence

from school climate empirical findings and current educational practice is socially unjust,

especially since research-based guidelines regarding positive youth development and

student learning have been established. Further, the researchers call for the need to close

the gap between school climate and school policy and practice to better support student

development and capacities for learning (Cohen et al., 2009; “Student-Centered High

Schools,” 2001).

Further school connectedness. Several exhaustive, multidisciplinary reviews

on the topic of school connectedness have attempted to adequately define school

connectedness and accurately measure the construct. The reviews revealed that the

variability of measures and the limited sampling of other studies calls for more research

to better establish the normative properties of a school connectedness measurement tool.

In addition, research would benefit from a wider, more diverse population of participants

to address sampling and reliability concerns (Cohen et al., 2009; Osterman, 2000;

Whitlock, 2006; Zullig et al., 2010).

Previous studies demonstrate that a connected school environment is related to

higher levels of student achievement. School connectedness, however, has been shown to

decrease through secondary school grade levels (Klem & Connell, 2004). Therefore, an

investigation of the relationship between the CHKS school connectedness variable and its

relationship to student achievement scores, especially at the seventh grade, may assist

educators with continued data-driven decisions related to school reform efforts and

illuminate the importance of school connectedness in order to sustain and support school

improvement efforts at the earliest secondary school grade (Zullig et al., 2010). The use

of the CHKS school connectedness scale as a survey instrument to explore the effects of

school connectedness as a mediating variable on the relationship between school

resilience measures, and the school academic achievement measure of API, would be the

first study of its kind.

Middle School Research

EdSource (2010) conducted a large-scale study of California’s middle grades,

which educates one of every eight students in Grades 6 through 8 in the United States,

meant to explore the relationship between school and district practices and policies and

the relationship to student achievement. The study cited the importance of middle grade

outcomes where research has shown that middle grades’ math performance is lower than

elementary grades. According to CDE data, two-thirds of fourth graders in California

scored proficient or advanced in mathematics in 2009, compared to only 43% of seventh

graders (Alspaugh, 1998; Anderman & Maehr, 1994; Picucci, Brownson, & Kahlert,

2002; Rich, 2005; Roesner & Eccles, 1998; Williams et al., 2010).

A review of middle grades literature from the past 20 years suggested that

Positive, Safe, and Engaging School Environments to be one of 10 important domains for

middle school study (Williams et al, 2010). The EdSource study concentrated on the

relationship between middle grades’ organizational practices and policies related to

improved CST outcomes in Language Arts and mathematics.

The study findings suggested that school environment was not associated with

improved student outcomes. This finding may perhaps be best explained by the fact that

California has no formal policy to encourage or mandate a research-based positive school

environment program in its general accountability system (Cohen et al., 2009). Thus, the

study did not find a strong association between California’s high-performing middle

school student outcomes and school environment policy. Other research, however, has

cited middle school climate as an important predictor of academic performance even after

controlling for student socioeconomic status (Hoy & Hannum, 1997). The findings

warrant continued middle school research and an investigation of the relationship

between school factors and student achievement.

Conclusion

The universal, generalist approach to resilience prepares students for school

success by enhancing environmental supports that address student health, development,

and well being. Further investigation regarding the relationship between school

protective factors (caring relationships, high expectation messages, and opportunities to

participate and contribute) and academic achievement should be explored.

Academic achievement as measured by standardized exams may offer limited

insights regarding positive school culture and youth development. Further investigation

may include an assessment of the underlying dimensions of school effectiveness:

instrumental and expressive functions (Eisner, 2001; Uline, Miller, & Tschannen-Moran,

1998).

CHAPTER 3—METHODOLOGY

Introduction

The purpose of this study was to investigate the relationship between student

resilience and academic achievement. This chapter presents the methodological

framework for the study, the research questions, instrumentation, the research design,

population, variables, and data analysis procedures.

This correlational study, with replicated procedures over three time points,

examined two data sets from multiple years. The Resilience Youth Development Module

(RYDM) of the California Healthy Kids Survey (CHKS) represents the first data set. The

second set of data is derived from the California Department of Education’s (CDE)

Academic Performance Index (API).

Most California schools administer the CHKS to meet the requirements of the

federal Safe and Drug Free Schools Communities Act (SDFSCA). Additionally, the

CDE identifies performance indicators that schools must monitor in meeting the

SDFSCA’s goals, under NCLB. The CHKS instrument helps schools monitor their goals

to maintain a safe and drug free school. An elementary school and secondary school

versions of the CHKS exists. This study concentrated specifically on the secondary

school, seventh grade responses.

The CHKS included a mandatory module administered to all students that focuses

on health behaviors and experiences (WestEd, 2006). Included within the CHKS

measure is the Resilience Youth Development Module (RYDM). The full RYDM

contains 56 items that were designed to measure internal student assets (personal

strengths) and external school resources (protective factors), all of which have been

linked to positive developmental outcomes (Benard & Slade, 2009).

The CHKS is an instrument that assesses school climate and culture. Research

suggests that bolstering student resilience positively impacts student achievement

(Benard, 1991). The API represents an aggregate school-wide score of student

achievement results. With this in mind, cross-sectional data sets of the CHKS and API

provided a base for the exploration of the notion that student resilience may be related to

student academic success. The following research questions directed this investigation:

Research Questions

1. Is there a significant statistical correlation between school protective factors of

caring relationships, high expectations, and meaningful participation to student

achievement?

2. Is there a predictive relationship between student internal assets of problem-

solving, self-efficacy, empathy, and self-awareness with student achievement?

3. Which protective factors and internal assets exhibit the most powerful

correlation with student achievement?

Instrumentation

External protective factors in school, such as caring relationships with teachers

and opportunities to participate, are recognized as protective factors; however, research in

this area is lacking (Sharkey et al., 2008). The absence of a psychometrically sound

instrument that reliably and validly measures the characteristics theorized to contribute to

student resilience in schools has limited the field of research (Jimerson, Sharkey, Nyborg,

& Furlong, 2004; Libby, 2004).

Despite several modifications to the RYDM, since its creation and initial

validation in 2000, the module lacked a thorough analysis of reliability and validity

(Furlong et al., 2009; Hanson & Kim, 2007). Recent empirical studies now support the

internal consistency and reliability of the psychometric properties of the RYDM. Hanson

and Kim (2007) conducted a detailed analysis of the RYDM’s survey items and found

that the number of items could be reduced, due to differential item functioning,

inconsistent factor loading patterns, or cross-loading across factors. Factor analyses

revealed that the survey items could generally be reduced from school Caring

Relationships and school High Expectation categories into one category: School Support.

The RYDM intended to measure student perceptions of school Caring Relationships and

school High Expectations as separate subscales. However, several factor analyses found

that the items from the two subscales could be combined into one scale, called School

Supports. School Meaningful Participation survey items, however, held together as a

separate factor (Hanson & Kim, 2007).

Hanson and Kim (2007) further suggested that certain items of the RYDM should

be dropped. Their research provided a rationale for dropping the items, ranging from

inconsistent functioning for certain ethnic populations, insufficient and invalid questions,

inconsistent functioning between genders, and cross loading in factor analyses (Furlong

et al., 2009). For example, a survey item related to “goals and plan for the future” and “I

plan to go to college or some other school after high school” functioned differently for

Chinese-American populations as compared to African American, Mexican American, or

White European American students. This difference suggested a survey item ethnic bias.

Therefore, the “goals and aspiration” and “cooperation” internal asset subscales were

deemed invalid leaving the constructs of self-efficacy, empathy, problem solving, and

self-awareness to be valid subscales.

Finally, Hanson and Kim (2007) examined RYDM item bias. They verified that

the factor structure of the scales held across racial-ethnic groups. Similarly, Furlong et al.

(2009) utilized Hanson and Kim’s factor structure to determine that the variance

attributable to grade (0.3%), ethnicity (0.8%), and gender (2.3%) were small. This study

provided normative data based on responses of 141,000 California students (Furlong

et al., 2009).

Lastly, further studies utilizing RYDM data provided empirical evidence that

reports of school assets and its relationship to individual resilience did not have a

differential relation when grouped by high- or low- self-reported levels of family (CFI =

0.980, NNFI = 0.977, and RMSEA = 0.033). That is, multigroup structural equation

modeling revealed that low family asset groups compared to high family asset groups

equally benefit from school assets (Sharkey et al., 2008).

The developing literature surrounding the RYDM and its normative properties, as

it related to ethnicity, gender, grade, and level of family assets, provides researchers the

potential of exploring the RYDM’s validated and reliable subscales and data. The

potential of applying the RYDM subscales to individuals in order to assess their potential

for academic success continues to develop (Furlong et al., 2009; Hanson & Austin, 2003;

Hanson & Kim, 2007; Sharkey et al., 2008). However, the validity of RYDM’s ability to

measure school-wide levels of resilience is substantiated. Given the research supporting

RYDM’s reliability and validity, and its widespread administration throughout the state

of California, the RYDM represents a psychometrically sound instrument that can

measure levels of student resilience in California schools.

This study utilized select-item CHKS RYDM items and subscales shown to be

reliable and valid in past studies (Appendix A).

Population

In 2004, the CDE mandated administration of the CHKS in all California schools.

This study used the 2004, 2006, and 2008 RYDM data from Grade 7 in all California

schools. In addition, this study used 2004, 2006, and 2008 CDE-calculated API scores

for all corresponding California schools. Schools with matched 2004, 2006, and 2008

CHKS data and API scores were collected, analyzed, and interpreted.

Research Design

The study examines extant data and incorporates a cross-sectional correlation

design that utilized a three-step statistical procedure for data analyses replicated over

three time periods: zero-order simple correlation; hierarchical multiple regression,

excluding the school connectedness variable; then, hierarchical multiple regression with

the school connectedness variable to test its effect as a mediator variable.

Individual CHKS survey-item responses (n = 1.5 million pupil cases) were

aggregated to produce school level responses. In turn, each school-level item response

was averaged, per resilience construct, to produce a composite score for each independent

variable related to resilience.

Schools with composite resilience scores, along with composite academic

achievement scores (API), were included in the study. In sum, 2004 data included

1,144 schools; 2006 data included 988 schools; and 2008 data included 837 schools.

Hierarchical multiple regression analysis illustrated the relationship between

survey data and achievement scores. Statistical analysis determined if a significant

statistical relationship between student survey data and student achievement existed.

Regression analysis controlled for student socioeconomic status, ethnicity, and other

demographic information. Further, hierarchical multiple regression analysis provided

evidence of the strength of the relationship between student perceptions and achievement

scores after other variables were accounted for. Standardized correlation coefficients

illustrated the relative strength of each resilience construct as it relates to student

achievement (Huck, 2008).

Limitations

Analysis of relationships using inferential analysis does not determine truths.

Statistical sampling and statistical analysis point to significant relationships between

events and the likelihood of occurring phenomena, but it cannot, with complete certainty,

directly attribute the occurrence of one event to another seemingly related occurrence

(Huck, 2008; Popham, 1993). Stated simply, the evidence of a correlation between

variables does not prove a causal relationship, nor does it indicate directionality. The

potential effect of another phenomenon, not explored as part of this study, may certainly

exist.

In addition, this study was limited by the sample, its measures, and its design.

The study was limited to 3 years of data in California seventh grade schools. Further,

the two measures utilized in this study (CHKS self-report survey and API) limited the

investigation of the resilience construct and its relationship to academic achievement.

Finally, the study was limited by a cross-sectional, 3-year correlational design that does

not utilize multiple measures, experimental manipulation, or random effects.

Variables

The independent variables of this study were the mean, composite scores of

each school’s student perceptions of nurturing adults, high expectations, meaningful

opportunities in school, as well as student self-reports of self-efficacy, empathy, problem

solving, and self-awareness. The external protective factors of nurturing adults and high

expectations were combined to a school support variable as suggested by previous

reliability and validity studies of the RYDM. In addition, the study included a school

connectedness variable from CHKS survey items to test its effect as a mediator variable.

The dependent variable is each school’s composite API score as calculated by the CDE.

The API factors schoolwide student achievement proficiencies in state-mandated

examinations.

Demographic predictor variables accounted for in the regression model included:

percentage of African-American students per school, percentage of Hispanic/Latino

children per school, percentage of English Language learners per school, and percentage

of students that were participants in the free or reduced price meal program per school.

Data Analysis Procedures

This study utilized both descriptive and inferential statistics. The one dependent

variable (API scores) and the multiple independent variables (RYDM of CHKS), along

with demographic variables, were analyzed through Hierarchical Multiple Regression

Analysis using the Statistical Package for the Social Sciences (SPSS), version 17.0.

Standardized correlation coefficients indicated the relative strength of the

relationship between the independent variables (resilience scores) and the dependent

variable (API scores). The probability of “F,” or the alpha-level, for each independent

variable to enter into the regression equation was set at .05.

A Hierarchical Multiple Regression Analysis revealed the predictive nature of the

RYDM survey responses in relation to API scores while testing the effects of school

connectedness as a mediator variable. The Multiple Regression also revealed which

independent variables most greatly affected API scores while accounting for the effects of

all other variables.

Finally, quantitative data analysis examined standardized and unstandardized

beta-weights to illustrate the relative strength of the relationship between the tested

variables (Huck, 2008).

Ethical Issues

This study posed no threat to teachers, students, or groups of students. Data

collected as part of this study are extant and already exist as aggregated school scores and

reports. All schools were coded to preserve the anonymity of school sites. Only the

relationship between schoolwide survey data and schoolwide academic performance are

reported. All data were kept secure according to the university’s Institutional Review

Board (IRB) guidelines.

CHAPTER 4—RESULTS

Introduction

The present study examined the relationship of school protective factors to student

achievement by utilizing composite select-item survey data from the California Healthy

Kids Survey (CHKS) and school composite scores from California’s Department of

Education, the Academic Performance Index (API). The chapter begins with a statistical

description of the data and the demographics. Then, a series of zero-order, simple

correlations presented by external school protective factors variables; student internal

assets variables; and school demographic variables follows. In addition, a school

connectedness variable was tested as a potential mediating variable; its findings conclude

the presentation of simple correlations. Next, a forced-entry, hierarchical regression

model tested the school connectedness variable as a potential mediator between other

variables; its findings present further correlation data after accounting for the effect of

each variable on the relationship to school API. Results of the multiple regression

models are presented by external school protective factors variables; student internal

assets variables; school demographic variables; and school connectedness variables.

Finally, the chapter concludes with a summary of results and findings to answer the three

research questions:

1. Is there a significant statistical correlation between school protective factors of

caring relationships, high expectations, and meaningful participation to student

achievement?

2. Is there a predictive relationship between student internal assets of problem-

solving, self-efficacy, empathy, and self-awareness with student achievement?

3. Which protective factors and internal assets exhibit the most powerful

correlation with student achievement?

Descriptive Statistics

The correlational, replication study utilized seventh grade CHKS data from 2003-

2004, 2005-2006, and 2007-2008 school years, along with corresponding school API

scores to illustrate the relationship between CHKS scores and API scores. Statistical

procedures were replicated in each of the three time periods. Each CHKS variable was

aggregated to a composite score. In each time period, schools with complete CHKS data

and an API score for the same year were included for study (n = 1,143 in 2004; n = 987 in

2006; n = 836 in 2008).

Academic Performance Index

The 2004 API mean score was 675.5 (SD = 119.7; range 281-953), the 2006 API

mean score was 687.7 (SD = 113.7; range 283-987), and the 2008 API mean score was

714.6 (SD = 113.1; range 332-982; Table 1).

Table 1

Academic Performance Index Descriptives by Year

2004 2006 2008

X variable Mean SD Mean SD Mean SD

API 675.49 119.70 687.74 113.62 714.59 113.14

External School Protective Factor Variables

External School Protective Factor Variables included School Support and

Meaningful Participation.

School Support. The 2004 School Support mean score was 2.9 (SD = .23; range

1.7-4.6), the 2006 School Support mean score was 2.8 (SD = .23; range 2.1-4.0), the 2008

School Support mean score was 3.0 (SD = .19; range 1.7-3.7; Table 2).

Table 2

External School Protective Factor Variables Descriptives by Year

2004 2006 2008

X variable Mean SD Mean SD Mean SD

Support 2.87 .23 2.83 .23 2.97 .19

Participation 2.28 .24 2.25 .23 2.26 .21

Meaningful Participation. The 2004 School Meaningful Participation mean

score was 2.3 (SD = .24; range 1.3-3.4), the 2006 School Meaningful Participation mean

score was 2.3 (SD = .23; range 1.3- 4.0), and the 2008 School Meaningful Participation

mean score was 2.3 (SD = .21; range 1.4-3.2; Table 2).

Internal Student Asset Variables

Internal student asset variables included Problem-Solving, Self-Efficacy,

Empathy, and Self-Awareness.

Problem-Solving. The 2004 Problem-Solving mean score was 2.8 (SD = .41;

range 1.0-4.0), the 2006 Problem-Solving mean score was 2.6 (SD = .55; range 1.0-4.0),

and the 2008 Problem-Solving mean score was 2.7 (SD = .57; range 1.0-4.0; Table 3).

Self-Efficacy. The 2004 Self-Efficacy mean score was 3.1 (SD = .51; range 1.0-

4.0), the 2006 Self-Efficacy mean score was 2.8 (SD = .67; range 1.0-4.0), and the 2008

Self-Efficacy mean score was 2.9 (SD = .70; range 1.0- 4.0; Table 3).

Table 3

Internal Student Asset Variables Descriptives by Year

2004 2006 2008

X variable Mean SD Mean SD Mean SD

Prob.-Solving 2.80 .41 2.59 .55 2.67 .57

Self-Efficacy 3.07 .51 2.79 .67 2.92 .70

Empathy 3.03 .48 2.76 .66 2.86 .68

Awareness 3.12 .46 2.85 .66 2.98 .68

Empathy. The 2004 Empathy mean score was 3.0 (SD = .48; range 1.0-4.0), the

2006 Empathy mean score was 2.8 (SD = .66; range 1.0-4.0), and the 2008 Empathy

mean score was 2.9 (SD = .68; range 1.0-4.0; Table 3).

Self-Awareness. The 2004 Self-Awareness mean score was 3.1 (SD = .46;

range 1.0-4.0), the 2006 Self-Awareness mean score was 2.9 (SD = .66; range 1.0-4.0),

and the Self-Awareness 2008 mean score was 3.0 (SD = .68; range 1.0-4.0; Table 3).

Demographic Variables

Demographics included percent African-American students, percent

Hispanic/Latino students, percent of students receiving free/reduced meals, and percent

English-learner students.

Percent African American. The 2004 mean percentage of African-American

students was 7.5 (SD = 11.6; range 0-87), 2006 mean percentage of African-American

students was 7.8 (SD = 11.3; range 0-83) and the 2008 mean percentage of African-

American students was 8.1 (SD = 10.8; range 0-81; Table 4).

Table 4

Demographic Variable Descriptives by Year

2004 2006 2008

X variable Mean SD Mean SD Mean SD

% African American 7.45 11.58 7.76 11.32 8.14 10.81

% Hispanic 36.63 26.90 43.31 28.46 42.90 27.47

% meals 39.83 26.67 45.09 28.02 45.82 28.10

% English- learners 16.24 16.03 18.79 16.55 18.71 15.45

Percent Hispanic/Latino. The 2004 mean percentage of Hispanic/Latino

students was 36.6 (SD = 26.9; range 0-99), 2006 mean percentage of Hispanic/Latino

students was 43.3 (SD = 28.5; range 0-100), and the 2008 mean percentage of Hispanic/

Latino students was 42.9 (SD = 27.4; range 0-100; Table 4).

Percent receiving meals. The 2004 mean percentage of students receiving free/

reduced school meals was 39.8 (SD = 26.7; range 0-100), the 2006 mean percentage of

students receiving free/reduced school meals was 45.1 (SD = 28.0; range 0-100), and the

2008 mean percentage of students receiving free/reduced school meals was 45.8 (SD =

28.1; range 0-100; Table 4).

Percent English-learners. The 2004 mean percentage of English-learner

students was 16.4 (SD = 16.0; range 0-94), the 2006 mean percentage of English-learner

students was 18.8 (SD = 16.6; range 0-98), and the 2006 mean percentage of English-

learner students was 18.7 (SD = 15.4; range 0-100; Table 4).

School Connectedness Variable

The 2004 School Connectedness mean score was 3.3 (SD = .31; range 1.7-4.6),

the 2006 School Connectedness mean score was 3.3 (SD = .32; range 1.8-5.0), and the

2008 School Connectedness mean score was 3.5 (SD = .29; range 2.2-4.5; Table 5).

Table 5

School Connectedness Descriptives by Year

2004 2006 2008

X variable Mean SD Mean SD Mean SD

Connectedness 3.32 .31 3.30 .32 3.49 .29

Simple Correlations With Academic Performance Index

The following section presents part 1 of a three-part statistical procedure for data

analysis. Mean scores of external school protective factors, internal student assets, school

demographic, and school connectedness variables were entered into a zero-ordered simple

correlation to school API scores for school years 2004, 2006, and 2008.

External School Protective Factor Variables

School Supports (caring relationships and high expectations) significantly

correlated with API in 2004 (r = +.28, p < .001), in 2006 (r = +.32, p < .001), and in 2008

(r = +.37, p < .001). These correlations were positive in direction, such that the higher the

support the higher the API scores (Table 6).

School Meaningful Participation significantly correlated with API in 2004

(r = +.43, p < .001), in 2006 (r = +.38, p < .001), and in 2008 (r = +.47, p < .001). These

correlations were positive in direction, such that the higher the participation, the higher

the API scores (Table 6).

Table 6

Correlations: External School Protective Factors and API by Year

Year Statistic School Support Meaningful Participation

AP104 r .28 .43

p <.001 <.001

N 1,143 1,143

AP106 r .32 .38

p <.001 <.001

N 987 987

AP108 r .37 .47

p <.001 <.001

N 836 836

These findings indicate that the external school protective factors of resilience,

School Supports (caring relationships and high expectations) and School Meaningful

Participation, are significantly correlated with student achievement scores, as represented

by API scores (Table 6).

Internal Student Asset Variables

Student problem-solving correlated with API in 2004 (r = +.22, p < .001), in 2006

(r = +.08, p = .012), and in 2008 (r = +.14, p < .001). These correlations were positive in

direction, such that the higher the levels of student problem-solving reports, the higher

the API score (Table 7).

Student self-efficacy significantly correlated with API in 2004 (r = +.19, p <

.001), in 2006 (r = +.08, p = .011), and in 2008 (r = +.11, p < .001). These correlations

Table 7

Correlations: Internal Student Assets and API by Year

Year Statistic Problem- Solving

Self- Efficacy Empathy Awareness

AP104 r .22 .19 .22 .19

p <.001 <.001 <.001 <.001

N 1,143 1,143 1,143 1,143

AP106 r .08 .08 .07 .04

p .012 .011 <.019 <.241

N 987 987 987 987

AP108 r .14 .11 .14 .08

p <.001 .001 <.001 .026

N 836 836 836 836

were positive in direction, such that the higher the levels of student self-efficacy, the

higher the API score (Table 7).

Student empathy significantly correlated with API in 2004 (r = +.22, p < .001), in

2006 (r = +.08, p = .019), and in 2008 (r = +.14, p < .001). These correlations were

positive in direction, such that the higher the levels of student empathy, the higher the

API score (Table 7).

Student self-awareness significantly correlated with API in 2004 (r = +.19, p <

.001) and in 2008 (r = +.08, p = .026). These correlations were positive in direction, such

that the higher the support, the higher the API scores. Student self-awareness in 2006

(r = +.04, p = .241), however, did not correlate with API score (Table 7).

Overall, these findings indicate that student internal assets of problem-solving,

self-efficacy, empathy, and self-awareness are significantly correlated with API scores.

School Demographic Variables

School percentages of African American students significantly correlated with

API in 2004 (r = -.33, p < .001), in 2006 (r = -.37, p < .001), and in 2008 (r = -.40,

p < .001). These correlations were negative in direction, such that the higher school

percentages of African American students, the lower the API scores (Table 8).

Table 8

Correlations: School Demographics and API by Year

Year Statistic % African American % Hispanic

% Reduced meals

% English learners

AP104 r -.33 -.50 -.51 -.40

p <.001 <.001 <.001 <.001

N 1,143 1,143 1,143 1,143

AP106 r -.36 -.49 -.56 -.43

p <.001 <.001 <.001 <.001

N 987 987 987 987

AP108 r -.40 -.51 -.56 -.44

p <.001 .001 <.001 <.001

N 836 836 836 836

School percentages of Hispanic/Latino students significantly correlated with API

in 2004 (r = -.50, p < .001), in 2006 (r = -.49, p < .001), and in 2008 (r = -.52, p < .001).

These correlations were negative in direction, such that the higher school percentages of

Hispanic/Latino students, the lower the API scores (Table 8).

School percentages of students qualifying for free/reduced meals significantly

correlated with API in 2004 (r = -.51, p < .001), in 2006 (r = -.56, p < .001), and in 2008

(r = -.56, p < .001). These correlations were negative in direction, suggesting that the

higher school percentages of students qualifying for free/reduced meals, the lower the

API score (Table 8).

School percentages of English-language learners significantly correlated with API

in 2004 (r = -.40, p < .001), in 2006 (r = -.43, p < .001), and in 2008 (r = -.44, p < .001).

These correlations were negative in direction, suggesting that the higher school

percentages of English-language learners, the lower the API score (Table 8).

Overall, these findings indicate that school demographics, such as: percentage of

African-American students, percentage of Hispanic/Latino students, percentage of

students receiving free/reduced meals, and percentage of English-language learners have

a significant inverse relationship with API scores. In other words, data indicate that

higher percentages of these student populations significantly correlate with lower API

scores (Table 8).

School Connectedness Variables

School Connectedness significantly correlated with API in 2004 (r = +.48,

p < .001), in 2006 (r = +.55, p < .001), and in 2008 (r = +.59, p < .001). These

correlations were positive in direction, such that the higher the levels of School

Connectedness, the higher the API score (Table 9).

Correlation Summary

A significant statistical correlation between external school protective factors of

school supports (caring relationships and high expectations) and school meaningful

participation with school API scores in 2004, 2006, and 2008 existed. Similarly, a

statistically significant, predictive relationship between levels of student internal assets,

such as: problem-solving, self-efficacy, empathy, and self-awareness with school API

Table 9

Correlations: School Connectedness and API by Year

Year Statistic School Connectedness

AP104 r .48

p <.001

N 1,143

AP106 r .55

p <.001

N 987

AP108 r .59

p <.001

N 836

scores existed in 2004, 2006, and 2008. In addition, the school connectedness variable

also significantly correlated with school API scores. In fact, school connectedness

exhibited the most positive predictive relationship with school API scores in 2004

(r = +.48, p < .001), in 2006 (r = +.55, p < .001), and in 2008 (r = +.59, p < .001;

Table 9). Further, each school demographic variable: percentage of African-American

students, percentage of Hispanic/Latino students, percentage of students receiving

free/reduced meals, and percentage of English-language learners had a significantly

negative inverse relationship with API scores in 2004, 2006, and 2008 (Table 8).

Hierarchical Multiple Regression

The following section presents parts 2 and 3 of a three-part statistical procedure

for data analysis. Correlation scores between external school protective factors, internal

student assets, and school demographic data variables to school API scores for school

years 2004, 2006, and 2008, were simultaneously entered into a regression equation to

account for the effect of all other variables in the relationship to school API. Then, a

forced-entry of the school connectedness variable into a second regression equation

accounting for all other variables tested its effect as a mediator variable.

External School Protective Factor Variables

School support was significantly predictive of API in 2004 (beta = -136.4,

p < .001), 2006 (beta = -90.8, p < .001), and 2008 (beta = -142.4, p < .001;

Appendices B-D) when school meaningful participation, problem-solving, self-efficacy,

empathy, self-awareness, percentage of African American students, percentage of

Hispanic/Latino students, percentage of students receiving free/reduced meals, percentage

of English- language learners, and school connectedness were accounted for. The

relationship was negative, such that the higher the support, the lower the API. School

support standardized betas for 2004 (beta = -0.257), 2006 (beta = -0.180), and 2008 (beta

= -0.236) indicate that each standard deviation increase in school support is related to a

decrease of roughly one-fifth of a standard deviation in API (Table 10). Combined, these

findings are not consistent with support as a positive predictor of API. Rather, support

was a significant negative predictor of API.

School meaningful participation was significantly predictive of API in 2004

(beta = +129.0, p < .001), 2006 (beta = +49.4, p = .003), and 2008 (beta = +84.1,

p < .001) when school supports, problem-solving, self-efficacy, empathy, self-awareness,

percentage of African American students, percentage of Hispanic/Latino students,

percentage of students receiving free/reduced meals, percentage of English-language

learners, and school connectedness were accounted for (Appendices B-D). The positive

Table 10

Coefficients With School Connectedness by Year

2004 2006 2008

X variable Standard

beta P Standard

beta P

Support -.257 <.001 -.180 <.001 -.236 <.001

Participation .254 <.001 .102 .003 .154 <.001

Problem solving .044 .274 .008 .850 .128 .027

Self-efficacy -.021 .677 .083 .118 -.195 .005

Empathy -.040 .424 .018 .728 .177 .015

Awareness .125 .005 -.136 .003 -.070 .230

% African- American -.163 <.001 -.191 <.001 -.203 <.001

% Hispanic/ Latino -.338 <.001 -.211 <.001 -.271 <.001

% meals -.261 <.001 -.277 <.001 -.176 <.001

% English- learners .092 .013 .006 .871 .009 .822

Connectedness .258 <.001 .404 <.001 .440 <.001

relationship indicated that the higher the school meaningful participation, the higher the

API. School meaningful participation standardized betas for 2004 (beta = +0.254), 2006

(beta = +0.102), and 2008 (beta = +0.154) indicate that each standard deviation increase

in school support is related to an increase of roughly one-sixth of a standard deviation in

API (Table 10). Combined, these findings are consistent with school meaningful

participation as a positive predictor of API.

Internal Student Asset Variables

Student problem-solving was not significantly predictive of API in 2004 (beta =

+12.9, p = .274) and 2006 (beta = +1.5, p = .850) when school supports, school

meaningful participation, self-efficacy, empathy, self-awareness, percentage of African

American students, percentage of Hispanic/Latino students, percentage of students

receiving free/reduced meals, percentage of English-language learners, and school

connectedness were accounted for (Appendices B and C). In 2008, however,

problem-solving (beta = +25.4, p = .027) was significantly predictive of API score after

accounting for the same variables listed above (Appendix D). The positive relationship

indicated that the higher the levels of student problem-solving, the higher the API.

Student problem-solving standardized betas for 2008 (beta = +0.128) indicate that

each standard deviation increase in school support is related to an increase of roughly

one-tenth of a standard deviation in API (Table 10). Combined, these findings do not

consistently illustrate that student problem-solving is a positive predictor of API.

Student self-efficacy was not significantly predictive of API in 2004 (beta = -4.9,

p = .677) and 2006 (beta = +14.1, p = .118) when school supports, school meaningful

participation, problem-solving, empathy, self-awareness, percentage of African American

students, percentage of Hispanic/Latino students, percentage of students receiving

free/reduced meals, percentage of English-language learners, and school connectedness

were accounted for (Appendices B and C). In 2008, however, self-efficacy (beta = -31.3,

p = .005) was significantly predictive of API score after accounting for the same variables

listed above (Appendix D). The relationship was negative, such that the higher the levels

of student self-efficacy, the lower the API. Student self-efficacy standardized betas for

2008 (beta = -0.195) indicate that each standard deviation increase in student self-efficacy

is related to a decrease of nearly one-fifth of a standard deviation in API (Table 10).

Combined, these findings are not consistent with student self-efficacy as a positive

predictor of API. Rather, student self-efficacy was a significant negative predictor of API

in 2008.

Student empathy was not significantly predictive of API in 2004 (beta = -10.0,

p = .424) and 2006 (beta = +3.0, p = .728) when school supports, school meaningful

participation, problem-solving, self-efficacy, self-awareness, percentage of African

American students, percentage of Hispanic/Latino students, percentage of students

receiving free/reduced meals, percentage of English-language learners, and school

connectedness were accounted for (Appendices B and C). In 2008, however, student

empathy (beta = +29.2, p = .015) was significantly predictive of API score after

accounting for the same variables listed above (Appendix D). The relationship was

positive, such that the higher the levels of student empathy, the higher the API. Student

empathy standardized beta for 2008 (beta = +0.177) indicate that each standard deviation

increase in student empathy is related to an increase of nearly one-fifth of a standard

deviation in API (Table 10). Combined, these findings illustrate that student empathy is

an inconsistent positive predictor of API.

Self-awareness (beta = -11.8, p = .230) was not significantly predictive of API

score in 2008, after accounting for school supports, school meaningful participation,

problem-solving, self-efficacy, empathy, percentage of African American students,

percentage of Hispanic/Latino students, percentage of students receiving free/reduced

meals, percentage of English-language learners, and school connectedness (Appendix D).

Student self-awareness, however, was significantly predictive of API in 2004 (beta =

+32.4, p = .005) and 2006 (beta = -23.6, p = .003) after accounting for the same variables

listed above (Appendix B and C). The positive relationship in 2004 indicated that the

higher the levels of student self-awareness, the higher the API. The negative relationship

in 2006 is such that the higher the levels of student self-awareness, the lower the API.

Student self-awareness standardized betas for 2004 (beta = +0.125) and 2006 (beta =

-0.136) contradictingly illustrate that a standard deviation increase in school support is

related to either an increase of roughly one-tenth of a standard deviation in API (2004) or

a decrease of roughly one-tenth of a standard deviation in API (2006) (Table 10).

Combined, these findings do not consistently illustrate that student self-awareness is as a

positive predictor of API.

School Demographic Variables

School percentage of African American students was significantly predictive of

API in 2004 (beta = -1.7, p < .001), 2006 (beta = -1.9, p < .001), and 2008 (beta = -2.1,

p < .001) when school support, school meaningful participation, problem-solving, self-

efficacy, empathy, self-awareness, percentage of Hispanic/Latino students, percentage of

students receiving free/reduced meals, percentage of English-language learners, and

school connectedness were accounted for (Appendices B–D). The relationship was

negative, such that the higher the percentage of African American students, the lower the

API. School percentage of African American students standardized betas for 2004

(beta = -0.163), 2006 (beta = -0.191), and 2008 (beta = -0.203) indicate that each standard

deviation increase in percentage of African American students is related to a decrease of

roughly one-fifth of a standard deviation in API (Table 10). Combined, these findings

suggest that percentage of African American students is a significant negative predictor of

API.

School percentage of Hispanic/Latino students was significantly predictive of API

in 2004 (beta = -1.5, p < .001), 2006 (beta = -0.8, p < .001), and 2008 (beta = -1.1,

p < .001) when school support, school meaningful participation, problem-solving, self-

efficacy, empathy, self-awareness, percentage of African American students, percentage

of students receiving free/reduced meals, percentage of English-language learners, and

school connectedness were accounted for (Appendices B–D). The relationship was

negative, such that the higher the percentage of Hispanic/Latino students, the lower the

API. School percentage of Hispanic/Latino students standardized betas for 2004 (beta =

-0.338), 2006 (beta = -0.211), and 2008 (beta = -0.271) indicate that each standard

deviation increase in percentage of Hispanic/Latino students is related to a decrease of

roughly one-fourth of a standard deviation in API (Table 10). Combined, these findings

suggest that percentage of Hispanic/Latino students is a significant negative predictor of

API.

School percentage of students receiving free/reduced meals was significantly

predictive of API in 2004 (beta = -1.2, p < .001), 2006 (beta = -1.1, p < .001), and 2008

(beta = -0.7, p < .001) when school support, school meaningful participation, problem-

solving, self-efficacy, empathy, self-awareness, percentage of African American students,

percentage of Hispanic/Latino students, percentage of English-language learners, and

school connectedness were accounted for (Appendices B–D). The relationship was

negative, such that the higher the percentage of students receiving free/reduced meals, the

lower the API. School percentage of students receiving free/reduced meals standardized

betas for 2004 (beta = -0.261), 2006 (beta = -0.277), and 2008 (beta = -0.176) indicate

that each standard deviation increase in percentage of students receiving free/reduced

meals is related to a decrease of roughly one-fifth of a standard deviation in API

(Table 10). Combined, these findings suggest that percentage of students receiving

free/reduced meals is a significant negative predictor of API.

School percentage of English-language learners was not significantly predictive of

API in 2006 (beta = +.043, p = .871) and 2008 (beta = +.064, p = .822) when school

supports, school meaningful participation, problem-solving, self-efficacy, empathy, self-

awareness, percentage of African American students, percentage of Hispanic/Latino

students, percentage of students receiving free/reduced meals, and school connectedness

were accounted for (Appendices C and D). In 2004, however, school percentage of

English-language learners (beta = +0.7, p = .013) was significantly predictive of API

score after accounting for the same variables listed above (Appendix B). The positive

relationship indicated that the higher the levels of school percentage of English-language

learners, the higher the API. School percentage of English-language learners

standardized betas for 2004 (beta = +0.092) indicate that each standard deviation increase

in school percentage of English-language learners is related to an increase of roughly

one-tenth of a standard deviation in API (Table 10). Combined, these findings do not

consistently illustrate that school percentage of English-language learners is as a positive

predictor of API.

School Connectedness Variable

School connectedness was significantly predictive of API in 2004 (beta = +100.0,

p < .001), 2006 (beta = +142.6, p = .001), and 2008 (beta = +173.0, p < .001) when

school supports, school meaningful participation, problem-solving, self-efficacy,

empathy, self-awareness, percentage of African American students, percentage of

Hispanic/Latino students, percentage of students receiving free/reduced meals, and

percentage of English-language learners were accounted for (Appendices B–D). The

positive relationship indicated that the higher the levels of school connectedness, the

higher the API. School connectedness standardized betas for 2004 (beta = +0.258), 2006

(beta = +0.404), and 2008 (beta = +0.440) indicate that each standard deviation increase

in school connectedness is related to an increase of roughly more than one-third of a

standard deviation in API (Table 10). Combined, these findings are consistent with

school connectedness as a positive predictor of API.

Hierarchical Multiple Regression Summary

The hierarchical multiple regression equation model accounted for variables, such

as: school supports, school meaningful participation, problem-solving, self-efficacy,

empathy, self-awareness, percentage of African American students, percentage of

Hispanic/Latino students, percentage of students receiving free/reduced meals, percentage

of English-language learners, and school connectedness.

After accounting for all other variables, the external school protective factor of

school supports was shown to be a significant negative predictor of school API, whereas

the external school protective factor of school meaningful participation was shown to be a

significant positive predictor of school API after accounting for the same variables. An

examination of internal student assets illustrated that each variable of problem-solving,

self-efficacy, empathy, and self-awareness do not consistently exhibit a significant

correlation to school API.

In contrast, demographic variables such as school percentage of African American

students, percentage of Hispanic/Latino students, and percentage of students receiving

free/reduced meals were shown to significantly, negatively correlate to school API after

accounting for all other variables. On the other hand, school percentage of English-

language learners did not consistently correlate to a predictive school API.

The most powerful correlation, positive or negative, after accounting for all other

variables is exhibited by the school connectedness variable. School connectedness had

the highest standardized betas for 2004 (beta = +0.258), 2006 (beta = +0.404), and 2008

(beta = +0.440) indicating that each standard deviation increase in school connectedness

is related to an increase of roughly more than one-third of a standard deviation in API

(Table 10). These findings illustrate that the school connectedness variable is the most

powerful predictor of school API.

School Connectedness Mediator Model

Simple correlation illustrated a statistically significant relationship between school

connectedness and school API (Table 9). However, when school connectedness was

added to the hierarchical multiple regression models, statistical analyses show that school

connectedness did not mediate the relationship between school protective factors and

school API, nor did it mediate the relationship between school internal assets and school

API. In other words, when the school connectedness variable was added to the regression

model, it did not significantly affect the relationship between the predictor variables and

the outcome variable; thus, it did not behave as a mediator variable (Appendices B–D).

Conclusion

The data from simple correlations and hierarchical multiple regressions between

the California Healthy Kids Survey (CHKS), in relation to California’s student

achievement composite score for schools, the Academic Performance Index (API),

allowed for the exploration of the relationship between school protective factors, student

internal assets, school demographics, and school connectedness to schoolwide student

achievement. Specifically, this study answered the following research questions:

1. Is there a significant statistical correlation between school protective factors of

caring relationships, high expectations, and meaningful participation to student

achievement?

Simple correlation data indicated a significant statistical correlation between

caring relationships and high expectations (school supports) along with school

meaningful participation to student achievement in 2004, 2006, and 2008. School

supports, however, did not significantly correlate consistently to student achievement

when exposed to other variables in multiple regression equations. On the other hand,

school meaningful participation consistently, positively correlated with student

achievement through 2004, 2006, and 2008, after all other variables were accounted for.

2. Is there a predictive relationship between student internal assets of problem-

solving, self-efficacy, empathy, and self-awareness with student achievement?

Overall, simple correlation data illustrated a consistently predictive positive

relationship between problem-solving, self-efficacy, empathy, and self-awareness with

student achievement in 2004, 2006, and 2008. However, when student internal asset

variables of problem-solving, self-efficacy, empathy, and self-awareness were entered

into a regression equation to account for the effect of other variables, none of the student

internal asset variables consistently exhibited a predictive relationship with student

achievement (API scores).

3. Which school protective factors and student internal assets exhibit the most

powerful correlation with student achievement?

After accounting for the effect of all school protective factor variables (school

supports and school meaningful participation) and student internal assets variables

(problem-solving, self-efficacy, empathy, and self-awareness) by entering them into the

equation that accounts for student demographic variables and a school connectedness

variable, school meaningful participation was shown to be one of the most consistent

positive predictors of student achievement. Even more powerful a predictor, however,

data indicated that the most powerful, positive predictor of student achievement was

shown to be the school connectedness variable. The school connectedness variable was

most predictive of the relationship with student achievement, not excluding the negative,

inverse relationship of student demographic data. As a point of comparison, the school

connectedness variable was nearly four times more powerful a predictor than school

meaningful participation in 2006 (Table 10), and nearly three-times more powerful a

positive predictor of student achievement scores than meaningful participation, the only

other consistently positive predictor variable of school API scores in 2008 (Table 10).

100

CHAPTER 5—SUMMARY, CONCLUSIONS,

AND RECOMMENDATIONS

Introduction

Academic success is more likely when teachers and administrators develop a

school culture that matches the developmental needs of adolescents (Comer, 2005).

Federal accountability measures and mandates, however, concentrate school reform

efforts on academic content standards measured in standardized tests. Unintended

consequences of NCLB legislation have resulted in a school leader test-focused

orientation and schools that have narrowed the curriculum, increased instructional pace,

created less engaging classrooms, and ignored student preferences for authentic, hands-on

learning, while attempting to raise test-score performance in order to avoid federal

sanctions (Certo et al., 2008). School leaders with a resilience-focus and orientation

towards providing a positive school environment, built upon positive school

relationships, can enhance a positive school culture, and thereby increase the potential

of improved academic outcomes (Bosworth & Earthman, 2002).

The purpose of this study was to explore the relationship between school

protective factors thought to promote student resilience and student academic

achievement. The study employed a correlational, three-step statistical procedure

replicated over three time periods utilizing extant select-item, self-report survey data from

seventh grade CHKS data in 2004, 2006, and 2008. The CHKS independent (predictor)

variables related to resilience were aggregated to school level scores, whereas the

dependent (outcome) variables were drawn from school level composite scores of school

API for the same years. The independent variables included the external school

101

protective factors of school supports and school meaningful participation, and the internal

student assets of problem-solving, self-efficacy, empathy, and self-awareness.

Statistical procedures for data analyses included simple correlation and

hierarchical multiple regressions, with and without school connectedness as a tested

mediator variable. The regression models accounted for school demographic variables

such as: percentage of African-American students per school, percentage of Hispanic/

Latino students per school, percentage of students receiving free/reduced school meals

per school, and percentage of English language learners per school.

The study answered the following research questions:

1. Is there a significant statistical correlation between school protective factors of

caring relationships, high expectations, and meaningful participation to student

achievement?

2. Is there a predictive relationship between student internal assets of problem-

solving, self-efficacy, empathy, and self-awareness with student achievement?

3. Which protective factors and internal assets exhibit the most powerful

correlation with student achievement?

This chapter discusses this study’s findings in relationship to empirical findings

and educational theory from previous studies. In the context of previous findings in

relationship to this study, the chapter goes further to discuss the study’s implications to

educational reform and leadership practice, and makes suggestions regarding how this

research may factor into educational decision-making. Finally, the chapter concludes

with a discussion of what limits this study and makes recommendations for areas of

102

educational leadership practice and for future research in order to promote and enhance

student academic achievement efforts.

Extension of Previous Research

The mixed results of this study did not provide any clear pattern to unambiguously

support all parts of the resilience construct and its relationship to improved academic

outcomes. This study, however, extends previous research in that a much larger sampling

of RYDM data was utilized over each of the three time periods when compared to the

previous studies and previous data sets (Hanson & Austin, 2003; Hanson et al. 2004).

The inclusion of the RYDM module to all CHKS administrations, beginning in 2004,

allowed for this study’s larger data set when compared to previous explorations.

In addition, previous studies that explored the relationship between CHKS and

RYDM survey data to academic outcomes utilized the results of the Standard

Achievement Test (SAT-9) standardized tests. The SAT-9 was the norm-referenced

standardized test used in California at the time of the previous studies (Hanson & Austin,

2003; Hanson et al., 2004). This study, however, updated the relationship between the

presently-used California standardized achievement tests, the criterion-referenced

California Standards Tests (CSTs), and select-item, psychometrically-sound CHKS

school resilience measures.

Further, the study drew parallels between the field of educational resilience and

educational leadership. The resilience construct suggested by the CHKS and RYDM was

derived from youth developmental models (Hanson & Austin, 2003), whereas the Social

Justice Leadership Theoretical Framework suggested by Theoharis (2009) stemmed from

educational leadership practice. Consequently, the study incorporated a unique

103

cross-discipline theoretical approach. Moreover, the study is the first to empirically test

the climate of the belonging component of the SJL framework by means of the school

connectedness variable of the CHKS, against an academic outcome.

Finally, this study contributed to school connectedness research in significant

ways. Previous school connectedness studies linked to academic outcome have been

limited to narrow populations lacking ethnic and economic diversity (Zullig et al., 2010).

The CHKS’ large data set of California seventh graders allowed for study of school

connectedness across a highly diverse, statewide sample of middle school students.

Summary of Findings

The following section summarizes the study’s findings regarding the relationship

between the external school protective factor variables, internal student assets variables,

and school connectedness variable to school API scores. The section concludes with an

explanation of how this empirical study extends previous research.

External School Protective Factors

Simple correlations between caring relationships and high expectations (school

supports) and school meaningful participation to school academic achievement were

shown to be statistically significant. This finding was consistent with past research

(Freiberg et al., 1995; Hanson & Austin, 2003; Hanson et al., 2004; Hawkins et al., 1999;

Wang et al., 1993; Waxman, Huang, & Padron, 1997; Waxman, Huang, & Wang, 1997).

Caring relationships with adults and high expectation messages, however, did not

consistently correlate to student achievement when entered into a regression equation that

accounted for other variables. This finding runs counter to other descriptive, correlational

studies that have explored the role of schools and its relationship to student achievement

104

which have suggested a correlation between caring adults and high expectations to

student achievement outcomes (Hanson & Austin, 2002, 2003; Hanson et al., 2004).

On the other hand, school meaningful participation was shown to be predictive of

increased academic outcome even after being exposed to other variables. Consistently

shown through three replicated procedures over three time periods, school meaningful

participation significantly correlated with higher API scores.

The CHKS RYDM construct suggesting that schools with caring adults, high

expectation messages, and meaningful participation positively correlates to increased

academic outcomes cannot be wholly validated by this study. The relationship between

external school protective factors and student academic outcome continues to prove to be

complex.

Internal Student Assets

The relationship between problem-solving, self-efficacy, empathy, and self-

awareness was shown to be statistically significant and predictive of outcome when

examined with simple correlations. Problem-solving, self-efficacy, empathy, and self-

awareness, however, were not predictive of student achievement when entered into a

regression equation that accounts for the effect of other variables.

The study findings do not support the CHKS RYDM model that suggests higher

levels of internal student assets are predictive of improved academic outcomes. The

CHKS RYDM resilience construct and its relationship to academic outcome cannot be

validated by this study.

105

School Connectedness

The school connectedness variable proved to exhibit the most powerful

correlation to student achievement. Simple correlations and hierarchical multiple

regressions consistently illustrated that school connectedness was a powerful predictor of

higher API scores even after all other variables were accounted for.

By comparison, school connectedness was three-to-four times more powerful a

predictor than school meaningful participation, the only other tested variable that was

shown to be statistically significant after simple correlations and multiple regressions

through 3 years of replicated statistical procedures.

The school connectedness variable is not part of the resilience construct. The

data, however, suggest a strong relationship between the constructs of school

connectedness, school meaningful participation, and student academic achievement.

Further resilience models may wish to consider the inclusion of the school connectedness

variable as part of the overall resilience construct.

Surprising Findings

School meaningful participation and school connectedness may be working in

tandem as variables that positively impact student achievement. These variables

subsumed the effects and contributions of other variables initially thought to have a

statistically significant relationship to student achievement. When exposed to the

regression equation with all other variables accounted for, all but school meaningful

participation and school connectedness demonstrated a weak correlation to student

achievement. A surprising finding of the study suggested that the caring, nurturing adult

106

and high expectation message variables were less-than-significant in relationship to

student achievement when all other variables were accounted for.

To add to unexpected findings, data from this study demonstrated no significant

relationship between student internal assets, thought to reflect student resilience, with

increased academic outcomes. These data may suggest a need to obtain other measures

of when and where students feel cared for, how and from whom they receive high

expectation messages, and of student internal assets of self-efficacy, empathy,

problem-solving, and self-awareness.

To examine the variables of caring adults and high expectations, for example,

aspects to consider may be the extent to which schools offer classes like band, music,

health, web design, the arts, peer counseling, leadership training, or other classes where

students feel as if they are positively contributing to the school, receiving a relevant

curriculum, and interacting with peers and teachers in classes that are interesting to them

and where they may receive feelings of success.

The test of school connectedness as a mediator variable gleaned unexpected

results. First, there was no evidence to suggest that school connectedness behaved as a

mediator variable. To determine if school connectedness functioned as a mediator

variable, two regression models were tested. The first regression model determined the

relationship between predictor variables and the outcome variables, without the inclusion

of the school connectedness variable. The second regression model examined the same

predictor variables and outcome variables, but included a school connectedness variable

within the regression model.

107

A comparison of the two regression models illustrated those relationships shown

to be statistically significant in the first model continued to be statistically significant

when the school connectedness variable was accounted for in the second regression

model. This finding suggests that school connectedness did not mediate the relationship

between predictor and outcome variables.

The power of the school connectedness variable, however, is noteworthy.

Consider the two hierarchical multiple regression models: one did not include the school

connectedness variable, while the other model included the school connectedness

variable. The difference between the models’ coefficients of determination illustrates the

power of the school connectedness variable. In 2004, 2006, and 2008, the school

connectedness variable, alone, accounted for respectively, 2%, 4%, and 6% (R² = .501,

.527, and .570, respectively) of the variance accounted for when determining the

variables’ relationship to student achievement.

Finally, the inclusion of the school connectedness variable in the multiple

regression equation resulted in a statistically significant relationship between school

supports and school API in 2006 and 2008. Initially, the relationship between school

supports and school API in 2006 and 2008 were not statistically significant. After adding

the school connectedness variable to the statistical model, a statistically significant

relationship between the school supports variable and school API was discovered. The

relationship, however, was negative in direction (2006 beta = -0.180 and 2008 beta =

-0.236) such that the higher the support, the lower the API.

The standardized betas indicated that each standard deviation increase in school

support is related to a decrease of roughly one-fifth of a standard deviation in school API.

108

This finding runs counter to the findings of previous resilience studies, and counter to the

notion of school supports relating to an increase of student academic outcomes (Freiberg

et al., 1995; Hanson & Austin, 2003; Hanson et al. 2004; Hawkins et al., 1999; Wang

et al., 1993; Waxman, Huang, & Padron, 1997; Waxman, Huang, & Wang, 1997).

The counter-intuitive finding begs the question of: why did the caring, nurturing

adults and high expectation messages variables result in a statistically significant negative

relationship, only after accounting for the school connectedness variable? The negative

relationship can perhaps be explained by the fact that schools with lower API scores may

treat students with higher levels of care and high expectations, but neglect to provide a

substantive curriculum that positively effects student outcome because the school is more

concentrated on the feelings and affect of the student, rather than providing an engaging

curriculum that enriches learning, school connectedness, or school meaningful

participation.

School leaders intent on optimizing school conditions for increased student

learning and achievement must foster teacher and student relationships grounded in

learning outcomes that promote school connectedness and school meaningful

participation. Efforts to nurture students and create caring relationships may

inadvertently excuse high expectations for teaching and learning by coddling students

and excluding them from a rigorous core curricula and relevant learning opportunities

because of poor test performance.

An examination of school master schedules may reveal institutionalized low

expectations when weighing the number of core courses versus remediation courses, for

example. In addition, rigor and high expectations imply more than just memorization and

109

test performance. Rigor implies measuring student abilities to demonstrate reasoning and

to apply the knowledge, relating to relevant learning (Wagner et al., 2006).

Educational change leadership theory would suggest that an effective teaching

model would include 3R’s: respectful relationships, rigorous core curricula, and relevant

curricula through real-world applications (Wagner et al., 2006). The 3R’s parallel the

external school protective factors of the resilience model, namely: caring relationships,

high expectations, and meaningful school participation.

General Discussion of Findings

The study findings revealed that there is a statistically significant, positive

relationship between external school protective factors of school meaningful participation

and school connectedness to student academic achievement. The data suggest that

schools with higher reports of school meaningful participation and higher reports of

school connectedness demonstrated higher student achievement scores. These current

findings support previous resilience study findings that demonstrate a relationship

between school environments that support and nurture student developmental needs and

increased student achievement outcomes (Freiberg et al., 1995; Hawkins et al., 1999;

Resnick et al., 1997; Rutter et al., 1979; Solomon, Battistich, et al., 1997; Solomon et al.,

2000; Solomon, Watson, et al., 1997; Wang et al., 1993; Waxman, Huang, & Wang,

1997; Wehlage et al., 1989).

Simple Correlations and Multiple Regressions

The presence of a significant relationship between student achievement and other

resilience construct variables of the resilience construct, such as: caring adults and high

expectation messages at school, along with student internal assets of self-efficacy,

110

empathy, problem-solving, and self-awareness were not verified by this study when all

variables were accounted for. In isolation, however, through simple correlations, the

resilience construct variables of caring adults and high expectation messages at school,

and student internal assets of self-efficacy, empathy, problem-solving, and self-awareness

were shown to be statistically significant and exhibited a relationship to student

achievement. Examining the correlational data in regression models, though, implied that

the effect and the contribution of the caring adult, high expectations, self-efficacy,

empathy, problem-solving, and self-awareness variables were subsumed by other

variables when all other variables were accounted for.

This finding demonstrated that simple and singular correlations of a set of

variables, alone, can be important, but may no longer prove to be statistically significant

when multiple variables are accounted for simultaneously. Examining the effects of

multiple variables, together, is particularly potent, and necessary, since the effect of

certain variables can be subsumed by other variables. In this study, for example, the

school connectedness and school meaningful participation variables subsumed the

contribution of other variables that were initially statistically significant.

This finding can also be explained by returning to the fundamental tenets of the

resilience model tested within this study. Recall that the model maintains students

develop internal assets of resilience to the degree their developmental needs are met

(Benard, 2004). According to study findings, the school meaningful participation mean

scores for students in 2004, 2006, and 2008 was 2.3 on a 4-point scale. This relatively

low mean score suggests that the average seventh grade survey participant in California

perceived that the presence of opportunities to: “do interesting activities in school, help

111

decide things like class activities or rules at school, and do things at school that make a

difference” was only “a little true.”

This study tested a model that included school meaningful participation as a key

developmental need that must be fulfilled if students are to build internal assets of student

resilience. Thus, it seems appropriate, to infer that since the existence of opportunities to

participate in meaningful ways was only “a little true” for the average student, subsequent

internal assets did not correlate with student achievement because the internal asset

variables, themselves, were underdeveloped (Benard, 2004).

To further explain why the internal student assets contribution was subsumed by

other variables, an examination of simple correlation data revealed that the magnitude

of the correlation between student internal assets and API scores to be very small and

insubstantial even before entering the regression models, and before being exposed to a

regression equation that accounts for all variables. Although the relationships were

shown to be statistically significant in 2008, for example, the correlation coefficients of

problem-solving, self-efficacy, empathy, and self-awareness with API were 0.14, 0.11,

0.14, and 0.08, respectively. The magnitudes of these correlation coefficients are

considered low and trivial, although statistically significant. The sheer number of cases

examined in this study contributed to the statistically significant relationship, but the

magnitudes of the correlations were weak (Huck, 2008; Popham, 1993). This finding

illustrates the need for educators to examine more than just statistically significant simple

correlations.

An examination of regression models, with multiple variables accounted for, may

provide educators with more telling data to better understand the relationship between

112

school variables and student achievement measures. Even more importantly, this finding

underscores how schools and school leaders should concentrate on the school protective

factors of meaningful school participation and school connectedness, shown to

demonstrate a moderate-to-major correlations to student achievement, even after all other

variables were accounted for.

Further Resilience Investigations

More exploration regarding resilience construct variables and its predictive

relationship to student achievement should be explored. Study findings suggest that

further investigation regarding the types of questions and instrumentation used to measure

the resilience construct may be needed. The anonymous nature of CHKS survey

collection prevents individual student-level resilience score correlations to student

achievement measures. Aggregated school-level resilience score correlations are limited

to generalized statistical inferences of the data. The entire field of educational resilience

would benefit from the continued use of a psychometrically sound school culture

measure, like the CHKS scales, to more broadly establish the normative properties of an

instrument.

In addition, this descriptive, correlation study explored a one-to-one relationship

between predictor variables and the outcome variable of API. Two or more variables may

be working together to affect the strength of other variables. A statistical analysis, such

as path analyses, may assist with identifying if a group of variables, in concert,

demonstrate a statistically significant relationship to student achievement.

It also seems plausible that other unknown factors and variables, not explored as

part of this study, are contributing to student performance outcomes. Students’ families,

113

communities, and peers, for example, are variables that are part of the resilience

construct, but not explored as part of this investigation. Further studies of the resilience

construct and its relationship to student achievement might include family, community,

and peers as other external protective factors.

Finally, the resilience model may benefit from the inclusion of a school

connectedness variable since findings from this study, other resilience studies, school

culture investigations, and educational leadership theories identify the close relationship

between school meaningful participation, school connectedness, and student achievement

(Whitlock, 2006). In sum, educational leaders with a resilience-focus, who consider

factors beyond test scores alone, account for the developmental needs of children while

monitoring school culture.

Implications for Educational Reform and Leadership Practice

In an age of NCLB, Blueprint for Reform, and Race to the Top accountability,

California educators continue to collect and use data to assess student needs and to

evaluate educational program efficacy and impact. The CHKS and RYDM measures are

often employed to stay within compliance of federal funding guidelines, but the resulting

data may be underutilized.

Educators and school leaders keen on creating optimized learning environments

may wish to aggregate CHKS school-level data to assess the extent to which students feel

as though their school provides meaningful ways to participate, and the extent to which

their students feel a sense of school connectedness since study findings suggested a strong

relationship to higher school API scores. An examination of the 2008 regression model

including school connectedness, utilizing standardized beta weights, for example, reveals

114

that a .29 increase to the school connectedness mean score of 3.49 is correlated to a

50-point increase in school API.

The study findings contribute to the field of educational leadership and school

reform inasmuch that it validates the use of the CHKS school connectedness and school

meaningful participation scales as measurements that positively correlate to school API.

The demonstration of a statistically significant relationship between the school

connectedness and school meaningful participation variables to school API points to the

strong relationship between school connectedness and student achievement, such that

higher reports of school connectedness, the higher school API. The findings provide

school leaders with compelling evidence that purposeful development of a school culture

that emphasizes school connectedness may positively impact schoolwide student

achievement.

Although only recently addressed through a growing number of school

connectedness studies, past studies illustrate that school connectedness levels can be

formally addressed to increase levels of school connectedness (Solomon, Battistich et al.,

1997). Despite the evidence that school connectedness can be fostered, and despite

investigations that demonstrate school connectedness has a relationship to student

academic outcomes, little educational policy exists to encourage the formal adoption of

programs that develop school levels of school connectedness (Osterman, 2000).

The study findings encourage school leaders to pursue formalized school

connectedness efforts, and administer CHKS surveys to a wide sampling of students in

order to measure levels of school connectedness through a diverse group of students with

varied experiences at school. Not only can the CHKS survey data be used formatively, to

115

identify levels of school connectedness that may be a key to improved school

performance, but summatively as well, to measure the extent to which adopted efforts

aimed at increasing school connectedness may be impacting levels of school

connectedness (Zullig et al., 2010).

School leaders that examine school connectedness and school meaningful

participation data may discover keys to increasing standardized test performance.

Further, the notion of the relationship between school connectedness and school

meaningful participation to increased student performance outcomes, may reasonably

be extended to potential increases in other standardized, criterion-referenced tests with

high-stakes implications, such as the California High School Exit Exam (CAHSEE), or to

other global measures of student performance, such as grades.

The findings from this study also suggest that, beyond looking at student

performance in state-tested core classes of math, English, the sciences, and social science,

school leaders should examine the data derived from school meaningful participation and

school connectedness in CHKS scores. The scores may be an effective measure to

examine the extent to which a school provides opportunities for meaningful school

participation and school connectedness in its curriculum, instruction, assessment, and

course offerings. For example, a school may concentrate on and purposefully attempt to

deliver an engaging, relevant curriculum with real-world applications while designing a

collaborative-team setting that enhances civic mindedness, social justice, equity, and

positive relationships. Successfully employed school connectedness and school

meaningful participation efforts would be reflected in higher reports of school

connectedness in CHKS data.

116

Study findings suggest delivering the curriculum within a learning context that

enhances school meaningful participation and school connectedness relate to a more

positive school culture and improved student achievement. The study suggests that a

focus on how instruction and curricula are delivered may be more important than

increased time-on-task in English or math, especially when attempting to spark student

learning, generate motivation, and sustain school engagement.

Limitations

The limitations of this study include limits to sample, measurement, and design.

Further, the study illustrates the relationship of the school survey data to student

achievement. The statistical sampling and statistical analysis help to illustrate a

relationship, but cannot attribute the occurrence of one event to another seemingly related

occurrence (Huck, 2008; Popham, 1993). In other words, the evidence of a correlation

between variables does not support a causal relationship, nor does it indicate

directionality.

Sample Limitations

The study is limited by the utilized sample. The sample was limited to schools

with 3 years of complete cross-sectional data in 2004, 2006, and 2008. In this case,

schools with complete data included those with an API score and CHKS data. In

addition, the study is limited by the findings from the California cohort of self-reporting

seventh grade student responses for each respective year. Future studies might replicate

the study design with other grade levels and other school culture and school resilience

scores across the nation.

117

Measurement Limitations

The study is further limited by the two instruments used to measure the

relationship between resilience scores and academic achievement. Namely, the construct

of academic achievement is solely measured by California’s Department of Education

composite score the Academic Performance index (API). Similarly, the construct of

resilience and school connectedness are solely measured by the results of the self-report

survey data gathered by the California Healthy Kids Survey (CHKS). The self-report

nature of the survey is a single measure that does not account for an objective measure or

an observable behavior measure. The use of one measure does not allow for a second

measure that may allow for convergence reliability of the resilience construct and its

relationship to student achievement.

Further, the use of data from all seventh grade California schools with resilience

scores matched to composite API scores limits the study to one grade level with no

treatment group. In turn, with no experimental treatment group, the study does not

account for random assignment or random effect when including all reporting California

schools to examine the relationship between resilience scores and student achievement.

Future studies might incorporate grades and student attendance as other student

achievement measures, and also consider qualitative methodology and focus groups to

account for observable measures.

Design Limitations

The study is also limited in its design. The examination of the relationship

between student resilience scores and student achievement is explored through a limited

sample and limited through the instruments used to measure the constructs. In addition,

118

the replicated, cross-sectional nature of the design does not account for a longitudinal

exploration of the data or its cohorts. For example, a longitudinal study of cohorts

through varying time points; such as, 5th grade, 7th grade, 9th grade, and 11th grade may

enrich the investigation as self-report survey information by age, through time, is

compared to student achievement scores.

Further, the study’s design does not account for variables, such as schools with

algebra and English support classes, types of remediation programs, incentive programs,

programs that intend on positively affecting student achievement, student attendance

rates, types of school culture, types of school leadership, or the school leadership’s

orientation towards building student resilience. It is conceivable that unknown and

unaccounted for variables may shed light on how schools can bolster academic

achievement.

Quantitative analysis of the variables that demonstrate a relationship to student

achievement can only be partially explained by a statistical model. A qualitative

investigation and approach to examining variables that are thought to be associated with

student achievement, however, such as the variables listed above, may lend insights to

understanding student, school, and school leader impacts on student achievement.

Varied Statistical Analyses

In addition to the potential of broadening and strengthening the findings of

this study by applying qualitative methodology, the study may also benefit from an

investigation of the external school protective factor variables, internal student asset

variables, school demographic variables, and the school connectedness variable, through

a varied statistical quantitative analysis. The statistical procedures of this study examined

119

12 distinct variables and each variable’s relationship to the student achievement outcome

measurement of school API through simple correlation and through hierarchical multiple

regressions that accounted for each variable’s effect on student achievement. The study

utilized a relatively large data set (n = 1.5 million student responses) that included data

from a wide sampling of California students aggregated to the school level and matched

to school level composite scores. The study findings were drawn from a large scale of

responses and the large scale of school API scores. The results were inclusive of all types

of schools.

Statistical manipulation of data, however, could potentially answer other research

questions not included as part of this study. For example, the data may have been

disaggregated by school type, ethnic composition, and base performance of the school.

A manipulation of data may provide evidence to answer questions such as: Did schools

with higher or lower API scores respond to the resilience measures differently? Would

differences in responses to levels of external school protective factors and internal

student assets vary by levels of school API scores? Would schools with higher or lower

percentages of students receiving free/reduced meals answer the external school

protective factors and internal student assets questions differently? Would an

investigation of schools with a certain percentage of demographic variables respond

differently when compared to other schools with a different ethnic or socioeconomic

composition? Would some groups of students appear to be more or less connected to

schools? Do certain types of schools foster more school meaningful opportunities to

participate?

120

Finally, this investigation examined survey responses of seventh grade students.

Would survey responses related to external school protective factors and internal student

assets vary by grade if examining 5th grade, 9th grade, or 11th grade responses? Findings

from these types of research questions may assist school leaders with the types of

programs employed as part of increasing student achievement.

Recommendations for Educational Practice

Social justice leadership (SJL) is one such educational leadership theory that

seems to incorporate school meaningful participation and school connectedness as part of

its operationalized construct. The SJL framework describes a three-legged approach to

improving schools, including: increased access to core learning, improved core learning,

and the creation of a climate of belonging.

The school meaningful participation and school connectedness variables, shown

to be powerfully consistent predictors of student achievement in this study, are addressed

within the SJL construct of increased access to core learning. Increasing access to core

learning presumes high expectations for all students and negates remedial pull-out

programs, where students are segregated from their peers to receive additional instruction

in tested basic skills.

A test-oriented focus on student achievement appears to adversely affect resilience

and youth development. The traditional pull-out, increased time-on-task model does not

account for student developmental needs to connect with school, and instead, results in

labeling, tracking, and a distancing from peers (Benard, 2004; Kohn, 2000; Meier, 2000;

Nieto, 1992; Oakes, 1985; Theoharis, 2009).

121

In addition, school meaningful participation and school connectedness are

fostered within the SJL construct of improved core learning. Educational equity

undergirds the SJL principle of improved core learning, which is, creating the sort of high

quality curricula, instruction, and assessments central to equitable classrooms. When

teachers strive for just and equitable learning opportunities for all students, students feel

that teachers care and hold high expectations for their success (Benard, 2004; Theoharis,

2009).

The notion of an equitable classroom is especially important when considering the

achievement gap between African-American and Hispanic/Latino students compared to

the achievements of White and Asian students when examining student achievement

measures. The findings of this study suggest that schools with higher percentages of

African-American students, Hispanic/Latino students, and students receiving free/reduced

meals, in general, possess lower API scores. These data suggest that lower performing

schools with higher percentages of African-American, Hispanic/Latino, and students

receiving free/reduced meals may benefit from instructional programs that promote

meaningful school participation, school connectedness, and SJL-oriented approaches.

Finally, SJL schools create a climate of belonging. This climate of belonging

parallels the construct of school connectedness, explored and substantiated within this

study. Social justice leadership bolsters school connectedness and a climate of belonging

by creating learning environments meant to engage students in collaborative learning

communities while incorporating social responsibility. Within SJL schools, students and

teachers exhibit a mutual respect for one another. In addition, a SJL school seeks to

engage students in designing their own learning activities that are interesting, allows a

122

student to help to decide class activities or rules, engages students in activities that are

relevant and make a difference, encourages school participation while treating all students

fairly, and promotes a sense of community (Theoharis, 2009).

Parallels between the resilience construct and the SJL-approach suggest that

resilience measures may help to determine the extent to which schools possess a SJL

orientation. As the findings of this study suggested, school levels of school meaningful

participation and school connectedness were associated with increased student

performance.

Utilizing an SJL approach, school leaders can enhance school resilience scores.

School leaders who recognize the relevant organizing construct of resilience may

approach school reform with a SJL, holistic and systems-based approach to school

organization, and may recognize that they can positively impact school culture and build

capacity for successful school change through SJL and resilience-focused school

programs (Bosworth & Earthman, 2002; Theoharis, 2009). Perhaps too often, however,

school leaders are focused solely on increasing test scores in response to external

pressures from the wider school community.

School Leader Orientation

Current NCLB legislation includes school and school district performance

mandates within its legislative construct. These mandates, however, meant to increase

the accountability of school leaders and to raise student achievement, pressure school

officials to adopt a test-focused orientation (Ravitch, 2010; Zhao, 2009).

Research suggests that school leaders, intent on organizational change and

successful school reform, should focus more broadly on aspects of positive school culture

123

and a systems-approach to student achievement (Fullan, 2000; Wagner et al., 2006). The

findings of this study suggest that attention be paid to student perceptions of school

meaningful participation and school connectedness to positively affect school

improvement efforts.

More resilience construct parallels can be drawn from Wagner et al.’s (2006)

Change Leadership construct that suggests school leaders and members of the school

community develop a common vision of effective teaching based on respectful, trusting

relationships, rigor, and relevance. Relevance of school curricula and instructional

strategies that explicitly develop the relevance of intended learning outcomes help to

develop meaningful participation and school connectedness (Wagner et al., 2006).

School leaders, therefore, should be mindful of resilience-focused constructs that

can positively enhance a school’s culture and a student’s level of school connectedness

and school meaningful participation, rather than pursuing test-focused outcomes and a

testing-orientation alone.

School Culture and School Reform

Deal and Peterson (2009) suggest that school leaders may often overlook the

importance of a positive and innovative organizational culture as a critical element to

school reform. They suggest that far too often, organizations are shaped by external

forces, such as NCLB mandates, rather than being shaped from within.

Given that NCLB mandates have failed to close the achievement gap and have

resulted in remediation programs that separate the lowest achieving students from

their peers, it seems appropriate that school leaders focus on acknowledging student

developmental needs to feel a sense of connectedness, belonging, and meaningful

124

participation to more positively affect school improvement and student achievement

efforts, and to build more positive school cultures (Cohen, 2006). The unintended

negative effects of a test-oriented society, demonstrated by China’s historically test-

driven educational system, coupled with the pervasive achievement gap in the United

States, should encourage school leaders to pursue sustainable reform efforts intent on

fostering a positive school culture and designed to engage students through a curriculum

that promotes school meaningful participation and school connectedness (Deal &

Peterson, 2009; Muhammed, 2009; Peters & Oliver, 2009; Ravitch, 2010; Zhao, 2009).

Student Achievement Variables

School leaders should consider student perceptions and accumulated CHKS

survey data to determine the extent to which students perceive their school provides

meaningful opportunities to participate and the extent to which students feel connected to

school. As an example, the most powerful predictor of increased schoolwide student

achievement outcomes was the school connectedness variable. Mean scores of school

connectedness on a 5-point scale through 2004, 2006, and 2008 were 3.2, 3.3, and 3.5,

respectively. The mean scores indicate that the average seventh grader in California

perceives that they neither agree, nor disagree, with the statements of: “I feel close to

people at this school; I am happy to be at this school; I feel like I am part of this school;

The teachers at this school treat students fairly; and I feel safe in my school.”

The study findings suggest that higher reports of school connectedness are related

to higher API scores. Given this information, school leaders should consider aspects of

the school’s culture and gathered student perceptions of school conditions for learning

which may be attained through various data, such as CHKS self-reports. Beyond

125

summative test performance data, and beyond remediation strategies that increase

instructional time, school leaders should consider how schools might bolster student

perceptions of school meaningful participation and school connectedness.

It seems incumbent on the school leader to survey school sites and to gather

various empirical data to most-accurately assess a school’s climate, capacity for school

improvement, and conditions for school change. A collection of varied data points

can assist school leaders with identifying school areas of need and may assist with

educational improvement efforts.

The replicated statistical models applied in this study, utilized over three time

points, provide significant support for the variables accounted for within this study. The

variables accounted for in the study: levels of external school protective factors of caring

adults, high expectations, meaningful opportunities for participation; levels of internal

student assets of self-efficacy, empathy, problem-solving, and self-awareness; school

demographic variables of percentage of African-American students, of percentage of

Hispanic/Latino students, of percentage of students receiving free/reduced student

lunches, and of percentage of English-language learners; and level of school

connectedness represented 50%, 53%, and 57% of the variance accounted for by the

hierarchical multiple regression model when determining the variables’ relationship to

student achievement (R² = .501, .527, and .570 in school years 2004, 2006, and 2008;

respectively). Stated simply, the study’s selected variables explain more than half of

possible variance when predicting the relationship between school factors and the student

achievement outcome measure of school API.

126

Accounting for more than half of the possible variance is significant, especially

when several other potential variables exist. To summarize, other variables not accounted

for in this study, but already suggested to be important factors for school leaders to

consider, are factors such as: the school leader’s perceptions towards school culture that

may reveal a test-focused or resilience-focused orientation; the extent to which the school

leader advocates for, and addresses student developmental needs; the extent to which the

school leader possesses an SJL orientation; the extent to which the school leader utilizes a

systems-based approach to school reform; the extent to which the school leader attributes

school culture as a component of school reform; or the extent to which the school leader

believes in developing positive relationships, collaboration, and community within the

school. Although not included as variables explored in this investigation, research and

educational theory suggests that the variables listed above are important to the decision-

making of a school leader and are related to student achievement outcomes (Bosworth &

Earthman, 2002; Deal & Peterson, 2009; Fullan, 2000; Muhammed, 2009; Ravitch, 2010;

Theoharis, 2009; Wagner et al., 2006; Zhao, 2009).

School Connectedness Obstacles

Research suggests that levels of school connectedness and levels of student

achievement decline with each subsequent year of secondary school (Whitlock, 2006).

School leaders continue to provide programs intended to remediate students performing

below grade-level standards and to improve student learning outcomes. Despite attempts

to improve the performance and learning of all students to address the well-documented

achievement gap, the gap persists.

127

It may be important to note the effect of the intervention programs and its

relationship to student perceptions. For example, are the programs that are meant to

support students perceived as they are intended? Does increased time-on-task in a subject

area that the student is already struggling with produce the intended outcome of student

achievement, or does it produce unintended consequences? Do students perceive being

supported or stifled; assisted or stigmatized; nurtured or suffocated; grouped by ability or

segregated? Do students receive a qualitatively different instructional program when

offered support classes, or do they get more of the same workbooks and exercises that

they have already tackled with little success? Do students recognize the relevance of the

subject matter with real-world applications? Do supported students sense that teachers

and schools have low expectations of them? Focus groups and student forums meant

to seek answers to these questions may help to direct school culture and school

improvement efforts. Rarely, however, are students asked for substantive feedback in

regards to their educational program (Hatchman & Rolland, 2001; Mitra, 2009; Whitlock,

2006).

Recommendations for Future Research

This study is limited to the examination of only two constructs: the measurements

of the CHKS survey and its resilience model and the student achievement measure of

school API. An attempt to triangulate the study’s findings with a qualitative exploration

to add convergent reliability may have been employed. For example, an intensely

concentrated small-scale study of schools with high and low resilience scores, high and

low API scores, with varying student demographic distributions utilizing focus group

128

interviews and observations could supplement and enrich the findings of this

investigation (Creswell, 2009).

In addition, replication of the statistical methods for children of other grade levels

such as 9th and 11th grades could have been compared with this study’s findings or

matched to different outcome variables, such as grades or another standardized test.

The use of multiple measures and experimental treatments, such as the application of a

resilience-focused program, may help to explain the complexities of the study’s findings,

the resilience measures, school settings, and student achievement.

Altogether, given that educators are continually working to improve student

outcomes, the need to explore new ways to affect change and meet the needs of diverse

student populations so that all students are performing at high levels is evident. Further

exploration may follow-up with the key variables shown to demonstrate a significantly

strong relationship to student academic achievement, namely: school connectedness and

school meaningful participation. An investigation of schools with higher reports of

school connectedness and school meaningful participation would inform this study.

An examination and utilization of other measures to explore variables that

positively impact student achievement would be an area of further research. Different

types of measures and more data would provide more convergence reliability to explore

the phenomenon of student achievement. This retrospective, descriptive study would

benefit from a prospective study that includes objective measures along with behavioral

measures. For example, further research may include other measures of school spirit,

school culture, or number and type of course offerings in the master schedule.

129

Further, survey data gathered from states beyond California that are designed to

measures constructs of student resilience, school connectedness, and school meaningful

participation may be utilized to examine the relationship to other student achievement

measures such as the Iowa’s Test of Basic Skills or Texas’ Assessment of Knowledge and

Skills. An analysis of the data from other types of survey data and its relationship to

varying student achievement measures may assist with providing normative data and

evidence to corroborate, or refute, the findings of this investigation.

Future study may explore the impact of programs and reforms meant to increase

student achievement. Social justice leadership suggests a promising construct for further

exploration. For example, common remediation efforts seek to increase a student’s

instructional time in math or English as a means to increase student achievement. Close

examination of the numbers and types of classes devoted to increased time on-task for

students struggling with core academic areas may prove illustrative, including

consideration of the way students are grouped throughout the school day. For example,

are struggling students homogeneously grouped by ability level, or is school intervention

incorporated within heterogeneously grouped classrooms?

In addition, further studies might explore the factors that contribute to school

connectedness and meaningful school participation. Investigations may include school

course offerings such as band, orchestra, art, sports programs, leadership class, and

programs designed to enhance community service, weighed against schools with fewer

such offerings and potentially more remediation courses that separate students by

achievement.

130

Finally, an investigation of schools that report higher levels of school

participation, school connectedness, and student achievement should be explored to

discover the mechanisms that foster such outcomes. A qualitative, ethnographic

approach may help reveal how schools that report higher levels of participation and

connectedness are developing higher levels of student achievement through the school’s

treatment of students and its approach to enhancing academic outcomes.

Conclusion

To summarize, the study showed a consistently significant statistical correlation

between the school meaningful participation variable and student achievement. Student

internal assets variables of problem-solving, self-efficacy, empathy, and self-awareness

did not demonstrate a predictive relationship to student achievement. And finally, the

school connectedness variable exhibited the most powerful correlation to student

achievement after all other variables were accounted for.

Improved student preparation and achievement continues to be a concern for

educators and our nation. Through further empirical investigations and a gathering of

diverse data points for analysis, educators will be able to make more informed decisions

regarding what schools can do to improve the achievement of all students.

131

REFERENCES

Alspaugh, J. (1998). Achievement loss associated with the transition to middle school and

high school. Journal of Educational Research, 92(1), 20-25.

Anderman, E., Austin, C., & Johnson, D. (2002). The development of goal orientation.

In A. Wigfield & J. Eccles (Eds.), Development of achievement motivation

(pp. 197-220). New York, NY: Academic Press

Anderman, E., & Maehr, M. (1994). Motivation and schooling in the middle grades.

Review of Educational Research, 64, 287-309.

Bandura, A. (Ed.). (1995). Self-efficacy in changing societies. Cambridge, UK:

Cambridge University Press.

Beardslee, W. (1997). Prevention and the clinical encounter. American Journal of

Orthropsychiatry, 68, 521-533.

Benard, B. (1991). Fostering resiliency in kids: Protective factors in the family, school,

and community. Portland, OR: Northwest Regional Educational Laboratory.

Benard, B. (2004). Resiliency: What we have learned. San Francisco, CA: WestEd.

Benard, B., & Marshall, K. (1997). A framework for practice: Tapping innate resilience.

Research/Practice, 5(1), 9-15.

Benard, B., & Slade, S. (2009). Listening to students: Moving from resilience research to

youth development practice and school connectedness. In R. Gilman, E. S.

Huebner, & M. J. Furlong (Eds.), Handbook of positive psychology in the schools

(pp. 353-370). New York, NY: Routledge.

Blum, R. W. (2005). A case for school connectedness. Educational Leadership, 67(7),

16-20.

132

Blum, R., Beuhring, T., Shew, M., Bearinger, L., Sieving, R., & Resnick, M., (2000). The

effects of race/ethnicity, income, and family structure on adolescent risk

behaviors. American Journal of Public Health, 90, 1879-1884.

Bosworth, K., & Earthman, E. (2002). From theory to practice: School leaders’

perspective on resiliency. Journal of Clinical Psychology, 58(3), 299-306.

Brown, J. H. (2004). Emerging social constructions in educational policy, research, and

practice. In H. C. Waxman, Y. N. Padron, & J. Gray (Eds.), Educational

resiliency: Student, teacher, and school perspectives (pp. 11-36). Charlotte, NC:

Information Age.

Brown, J. H., & D’Emidio-Caston, M. (1995). On becoming at-risk through drug

education: How symbolic policies and their practices affect students. Evaluation

Review, 19(4), 451-492.

Brown, J. H., D’Emidio-Caston, M., & Pollard, J. (1997). Student and substances: Social

power in drug education. Educational Evaluation and Policy Analysis, 19(1),

65-82.

Brownlee, S. (1996, November 11). Invincible kids. U.S. News & World Report, 62-71.

Calabrese, R. L., Hummel, C., & San Martin, T. (2007). Learning to appreciate at-risk

students. International Journal of Educational Management, 21, 275-291.

California Department of Education (CDE). (2009). 2008-2009 Academic performance

index reports. Retrieved from http://www.cde.ca.gov/ta/ac/ap/documents/

infoguide08.pdf

Catterall, J. (1998). Risk and resilience in student transitions to high school. American

Journal of Education, 106, 302-333.

133

Cefai, C. (2008). Promoting resilience in the classroom: A guide to developing pupils’

emotional and cognitive skills. London, England: Kingsley Publishers.

Certo, J. L., Cauley, K. M., & Moxley, K. D. (2008). An argument for authenticity:

Adolescents’ perspectives on standards-based reform. The High School Journal,

91, 26-38.

Cohen, J. (2006). Social, emotional, ethical, and academic education: Creating a climate

for learning, participation in democracy, and well-being. Harvard Educational

Review, 76(2), 201-285.

Cohen, J., McCabe, E. M., Michelli, N. M., & Pickeral, T. (2009). School climate:

Research, policy, practice, and teacher education. Teachers College Record,

111(1), 180-213.

Comer, J. P. (2005). Child and adolescent development: The critical missing focus in

school reform. Phi Delta Kappan, 86(10), 757-763.

Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed methods

approaches. Thousand Oaks, CA: Sage.

Criss, M., Petit, G., Bates, J., Dodge, K., & Lapp, A. (2002). Family adversity, positive

peer relationships, children’s externalizing behavior: A longitudinal perspective

on risk and resilience. Child Development, 73(4), 1220-1237.

Csikszentmihalyi, M. (1990). Flow: The psychology of optimal experience. New York,

NY: HarperCollins.

Cuban, L. (2007). Hugging the middle: Teaching in an era of testing and accountability,

1980-2005. Education Policy Analysis Archives, 15(1), 1-29.

134

Deal, T. E., & Peterson, K. D. (2009). Shaping school culture: Pitfalls, paradoxes, and

promises. San Francisco, CA: Jossey-Bass.

Deci, E. (1995). Why we do what we do: Understanding self-motivation. New York, NY:

Penguin Books.

Eccles, J., & Gootman, J. (2002). Community programs to promote youth development.

Washington, DC: National Academic Press.

Eisner, E. W. (2001). What does it mean to say a school is doing well? Phi Delta

Kappan, 82, 367-372.

Elias, M. J. (2001, Winter). Middle school transition: It’s harder than you think: Making

the transition to middle school successful. Middle Matters, 1-2.

Farmer, T., Estell, D., Leung, M., Trott, H., Bishop, J., & Cairns, B. (2003). Individual

characteristics, early adolescent peer affiliations, and school dropout: An

examination of aggressive and popular group types. Journal of School

Psychology, 41, 217-232.

Florida, R. (2005). The flight of the creative class: The new global competition for talent.

New York, NY: Harper Business.

Freiberg, H., Stein, T., & Huang, S. (1995). The effect of classroom management

intervention on student achievement in inner city elementary schools. Educational

Research and Evaluation, 1(1), 33-66.

Freire, P. (1993). Education for critical consciousness. New York, NY: Continuum.

Fowler, F. C. (2009). Policy studies for educational leaders: An introduction. Boston,

MA: Pearson.

135

Fullan, M. (2000). The three stories of educational reform: Inside; inside/out; outside/in.

Phi Delta Kappan, 81, 581-584.

Fullan, M. (2001). Leading in a culture of change. San Francisco, CA: Jossey-Bass.

Furlong, M. J., Ritchey, K. M., & O’Brennan L. M. (2009). Developing norms for the

California Resilience Youth Development Module: Internal assets and school

resources subscales. California School Psychologist, 14, 35-46.

Gardner, H. (1993). Multiple intelligences: The theory in practice. New York, NY: Basic

Books.

Garmezy, N., Masten, A. S., & Tellegen, A. (1984). The study of stress and competence

in children: A building block of developmental psychopathology. Child

Development, 55, 97-111.

Goleman, D. (1995). Emotional intelligence: Why it can matter more than I.Q. New

York, NY: Bantam Books.

Gonzalez, R., & Padilla, A. M. (1997). The academic resilience of Mexican-American

high school students. Hispanic Journal of Behavioral Sciences, 19, 301-317.

Hagborg, W. J. (1994). An exploration of school membership among middle- and high-

school students. Journal of Psychoeducational Assessment, 12, 313-323.

Hanson, T. L., & Austin, G. A. (2002). Health risks, resilience, and the Academic

Performance Index. (California Healthy Kids Survey Factsheet 1). Los Alamitos,

CA: WestEd.

Hanson, T. L., & Austin, G. A. (2003). Student health risks, resilience, and academic

performance in California: Year 2 report, longitudinal analyses. Los Alamitos,

CA: WestEd.

136

Hanson, T. L., Austin, G. A., & Lee-Bayha, J. (2004). Ensuring that no child is left

behind: How are student health risks & resilience related to the academic

progress of schools? San Francisco, CA: WestEd.

Hanson, T. L., & Kim, J. O. (2007). Measuring resilience and youth development: The

psychometric properties of the Healthy Kids Survey. (Issues & Answers Report,

REL 2007–No. 034). Washington, DC: U.S. Department of Education, Institute of

Education Sciences, National Center for Education Evaluation and Regional

Assistance, Regional Educational Laboratory West.

Hatchman, J. A., & Rolland, C. (2001, April). Students’ voices about schooling: What

works for them—It’s implications to school reform. Paper presented at the annual

meeting of the American Educational Research Association, Seattle, WA.

Hawkins, J., Catalano, R., Kosterman, R., Abbot, R., & Hill, K. G. (1999). Preventing

adolescent risk behaviors by strengthening protection during childhood. Archives

of Pediatrics and Adolescent Medicine, 153, 226-234.

Hawkins, J. D., Catalano, R. F., & Miller, J. Y. (1992). Risk and protective factors for

alcohol and other drug problems in adolescence and early adulthood: Implications

for substance abuse prevention. Psychological Bulletin, 112(1), 64-105.

Henderson, N., & Millstein, M. (1992). Resiliency in schools. Thousand Oaks, CA:

Corwin Press.

Higgins, G. (1994). Resilient adults: Overcoming a cruel past. San Francisco, CA:

Jossey-Bass.

137

Hoy, W. K., & Hannum, J. W. (1997). Middle school climate: An empirical assessment

of organizational health and student achievement. Educational Administration

Quarterly, 33, 290-311.

Huck, S. (2008). Reading statistics and research (5th ed.). Boston, MA: Pearson

Education.

Hupfeld, K. (2007). Resilience skills and dropout prevention: A review of the literature.

Retrieved from http://scholarcentic.com/images/pdf/resiliency_skills/SC_

Resiliency_WP_FNL.pdf

Jimerson, S. R., Sharkey, J. D., Nyborg, V. M., & Furlong, M. J. (2004). Strength-based

assessment and school psychology: A summary and synthesis. California School

Psychologist, 9, 9-20.

Kaufman, J., Cook, A., Arny, L., Jones, B., & Pittinsky T. (1994). Problems defining

resilience: Illustrations from the study of maltreated children. Development and

Psychopathology, 6, 215-229.

Klem, A. M., & Connell, J. P. (2004). Relationships matter: Linking teacher support to

student engagement and achievement. Journal of School Health, 74(7), 262-273.

Kohn, A. (2000). The case against standardized testing: Raising the scores, ruining the

schools. Portsmouth, NH: Heinmann.

Larson, R. (2000). Toward a psychology of positive youth development. American

Psychologist, 55, 170-183.

138

Lepper, M., Sethi, S., Dialdin, D., & Drake, M. (1997). Intrinsic and extrinsic motivation:

A developmental perspective. In S. Luthar, J. Burack, D. Cicchetti, & J. Weisz

(Eds.), Developmental psychopathology: Perspectives on adjustment, risk, and

disorder (pp. 23-50). New York, NY: Cambridge University Press.

Libby, H. P. (2004). Measuring student relationships to school: Attachment, bonding,

connectedness, and engagement. Journal of School Health, 74, 274-283.

Liddle, H. (1994). Contextualizing resilience. In M. Wang & E. Gordon (Eds.),

Educational resilience in inner-city America: Challenges and prospects

(pp. 45-72). Hillsdale, NJ: Erlbaum.

Luthar, S. (2003). Resilience and vulnerability: Adaptation in the context of childhood

adversities. Cambridge, UK: Cambridge University Press.

Luthar, S., & Burak, J. (2000). Adolescent wellness: In the eye of the beholder? In

D. Cicchetti, J. Rappaport, I. Sandler, & R. Weissberg (Eds.), The promotion of

wellness in children and adolescents (pp. 29-57). Washington, DC: Child Welfare

League Association Press.

Luthar, S., Cicchetti, D., & Becker, B. (2000). The construct of resilience: A critical

evaluation and guidelines for future work. Child Development, 71(3), 543-562.

Masten, A. (1994). Resilience in individual development: Successful adaptation despite

risk and adversity. In M. Wang & E. Gorden (Eds.), Educational resilience in

inner-city America (pp. 3-25). Hillsdale, NJ: Lawrence Erlbaum.

Masten, A. (2001). Ordinary magic: Resilience processes in development. American

Psychologist, 56, 227-238.

139

Masten, A., & Coatsworth, D. (1998). The development of competence in favorable and

unfavorable environments: Lessons from research on successful children.

American Psychologist, 53, 205-220.

McClendon, C., Nettles, S. M., & Wigfield, A. (2000). Fostering resilience in high school

classrooms: A study of the PASS Program (Promoting Achievement In School

Through Sport). In M. G. Sanders (Ed.), Schooling students placed at risk:

Research, policy, and practice in the education of poor minority adolescents

(pp. 289-307). Mahwah, NJ: Lawrence Erlbaum.

Meier, D. (2000). Will standards save public education? Boston, MA: Beacon Press.

Miller, A. (1990). The untouched key: Tracing childhood trauma in creativity. New York,

NY: Anchor Books.

Mitra, D. L. (2009). Strengthening student voice initiatives in high schools: An

examination of the supports needed for school-based youth-adult partnerships.

Youth & Society, 40, 311-335.

Muhammad, A. (2009). Transforming school culture: How to overcome staff division.

Bloomington, IN: Solution Tree Press.

National Commission on Excellence in Education. (1983). A nation at risk: The

imperative for educational reform. Washington, DC: Government Printing Office.

Nettles, S. M., Mucherah, W., & Jones, D. S. (2000). Understanding resilience: The role

of social resources. Journal of Education for Students Placed At Risk, 5, 47-60.

Nieto, S. (1992). Affirming diversity: The sociopolitical context of multicultural

education. New York, NY: Longman.

140

Oakes, J. (1985). Keeping track: How students structure inequality. New Haven, NJ:

Yale University Press.

Osterman, K. F. (2000). Students’ need for belonging in the school community. Review of

Educational Research, 70, 323-367.

Padron, Y., Waxman, H., & Huang, S. (1999). Classroom and instructional learning:

Environment differences between resilient and nonresilient elementary school

students. Journal of Education for Students Placed At Risk, 4(1), 63-81.

Padron, Y. N., Waxman, H. C., Powers, R. A., & Brown, A. (2002). Evaluating the

effects of pedagogy to improve resiliency programs on English language learners.

In L. Minaya-Rowe (Ed.), Teacher training and effective pedagogy in the context

of student adversity (pp. 211-238). Greenwich, CT: Information Age.

Peters, S., & Oliver, L. A. (2009). Achieving quality and equity through inclusive

education in an era of high-stakes testing. Prospects, 39, 265-279.

Picucci, A. C., Brownson, A., & Kahlert, R. (2002). Driven to succeed: High-performing,

high-poverty, turnaround middle schools. Austin, TX: University of Texas at

Austin, Charles A. Dana Center.

Popham, W. J. (1993). Educational evaluation. Needham Heights, MA: Allyn & Bacon.

Popham, W. J. (2001). The truth about testing: An educator’s call to action. Alexandria,

VA: Association for Supervision and Curriculum Development.

Ravitch, D. (2010). The death and life of the great American school system: How testing

and choice are undermining education. Philadelphia, PA: Basic Books.

141

Resnick, M., Bearman, P., Blum, R., Bauman, K., Harris, L., & Jones, J. (1997).

Protecting adolescents from harm: Findings from the national longitudinal study

on health. Journal of the American Medical Association, 278, 823-832.

Rhodes, W. A., & Brown, W. K. (Eds.). (1991). Why some children succeed despite the

odds. New York, NY: Praeger.

Rich, N. (2005). Motivating at-risk middle school students to positive classroom

performance. ERS Spectrum, 23, 23-31.

Roderick, M., Chiong, J., DaCosta, K., Arney, M., Stone, S., & Waxman E. (1997). Final

report 1: Sample recruitment, composition, and retention methods and results.

Chicago, IL: University of Chicago.

Roeser, R. W., & Eccles, J. (1998). Adolescents’ perceptions of middle school: Relation

to longitudinal changes in academic and psychological adjustment. Journal of

Research on Adolescence, 8, 123-158.

Rutter, M. (1987). Psychosocial resilience and protective mechanism. American Journal

of Orthopsychiatry, 57, 316-331.

Rutter, M. (1989). Pathways from childhood to adult life. Journal of Child Psychology

and Psychiatry, 30, 23-54.

Rutter, M. (2000). Resilience reconsidered: Conceptual considerations, empirical

findings, and policy implications. In J. P. Shonkoff & S. J. Meisels (Eds.),

Handbook of early childhood intervention (2nd ed., pp. 651-682). New York, NY:

Cambridge University Press.

Rutter, M., Maughn, B., Mortimore, P., & Ouston, J. (1979). Fifteen thousand hours:

Secondary schools and their effects on children. London, England: Open Books.

142

Scales, P., & Leffert, N. (1999). Developmental assets: A synthesis of the scientific

research on adolescent development. Minneapolis, MN: Search Institute.

Sharkey, J. D., You, S., & Schnoebelen, K. (2008). Relations among school assets,

individual resilience, and student engagement for youth grouped by level of family

functioning. Psychology in the Schools, 45, 402-418.

Solomon, D., Battistich, V., Il-Kim, D., & Watson, M. (1997). Teacher practices

associated with students’ sense of the classroom as a community. Social

Psychology of Education, 1, 235-267.

Solomon, D., Battistich, V., Watson, M., Schaps, E., & Lewis, C. (2000). A six-district

study of educational change: Direct and mediated effects of the child development

project. Social Psychology of Education, 4, 3-51.

Solomon, D., Watson, M., Battistich, V., Schaps, E., & Delucchi, K. (1997). Creating

classrooms that students experience as communities. American Journal of

Community Psychology, 24(6), 719-748.

Stohlberg, A., & Mahler, J. (1994). Enhancing treatment gains in a school-based

intervention for children of divorce through skill training, parental involvement,

and transfer procedures. Journal of Consulting and Clinical Psychology, 62,

147-156.

Student-centered high schools: Helping schools adapt to the learning needs of

adolescents. (2001, September). Perspectives on Policy and Practice. Providence,

RI: Brown University, Northeast and Islands Regional Educational Laboratory.

Theoharis, G. (2009). School leaders our children deserve: Seven keys to equity, social

justice, and school reform. New York, NY: Teachers College Press.

143

Uline, C., Miller, D., & Tschannen-Moran, M. (1998). School effectiveness: The

underlying dimensions. Educational Administration Quarterly, 34(4), 462-483.

U.S. Department of Education (USDOE). (2002). The No Child Left Behind Act of 2001.

Retrieved from http://www2.ed.gov/nclb/overview/intro/execsumm.pdf

U.S. Department of Education (USDOE). (2010). A blueprint for reform: Reauthorization

of the Elementary and Secondary Education Act. Retrieved from

http://www2.ed.gov/ policy/elsec/leg/blueprint/index.html

Wagner, T., Kegan, R., Lahey, L., Lemons, R. W., Garnier, J., Helsing, D., . . .

Rasmussen, H. T. (2006). Change leadership: A practical guide to transforming

our schools. San Francisco, CA: Jossey-Bass.

Wang, M. C., Haertel, G. D., & Wahlberg, H. J. (1993). Synthesis of research: What

helps students learn? Educational Leadership, 51(4), 74-79.

Wang, M. C., Haertel, G. D., & Wahlberg, H. J. (1994). Educational resilience in inner

cities. In M. C. Wang & E. W. Gordon (Eds.), Educational resilience in inner-city

America: Challenges and prospects (pp. 45-72). Hillsdale, NJ: Erlbaum.

Wang, M. C., & Reynolds, M. C. (Eds.). (1995). Making a difference for students at risk:

Trends and alternatives. Thousand Oaks, CA: Corwin.

Watt, N., David, J., Ladd, K., & Shamos, S. (1995). The lifecourse of psychological

resilience: A phenomological perpective of deflecting life’s slings and arrows.

Journal of Primary Prevention, 15, 209-246.

144

Waxman, H., Brown, A., & Chang, H. (2004). Future directions for educational resiliency

research. In H. Waxman, Y. Padron, & J. Gray (Eds.), Educational resiliency:

Student, teacher, and school perspectives (pp. 263-273). Greenwich, CT:

Information Age Publishing.

Waxman, H. C., Gray, J. P., & Padron, Y. N. (2003). Review of research on educational

resilience. Santa Cruz, CA: University of California, Center for Research on

Education, Diversity, and Excellence. Retrieved from http//www.escholarship.

org/uc/item/7x695885

Waxman, H., & Huang, S. (1996). Motivation and learning environment differences

between resilient and nonresilient inner-city middle school students. Journal of

Educational Research, 90, 93-102.

Waxman, H., Huang, S., & Padron, M. (1997). Motivation and learning environment

differences between resilient and nonresilient Latino middle school students.

Hispanic Journal of Behavioral Sciences, 19, 137-155.

Waxman, H., Huang, S., & Wang, M. (1997). Investigating the multilevel classroom

learning environment of resilient and nonresilient students from inner-city

elementary schools. International Journal of Educational Research, 27, 343-353.

Wehlage, G., Rutter, R., Smith, G., Lesko, N., & Fernandez, R. (1989). Reducing the

risk: Schools as communities of support. London, England: Falmer Press.

Werner, E. (1986). Resilient offspring of alcoholics: A longitudinal study from birth to

age 18. Journal of Studies on Alcohol, 14, 34-40.

Werner, E., & Smith, R. (1982). Vulnerable but invincible: A longitudinal study of

resilient children and youth. New York, NY: McGraw-Hill.

145

Werner, E., & Smith, R. (1992). Overcoming the odds: High risk children from birth to

childhood. New York, NY: Cornell University Press.

Werner, E., & Smith, R. (2001). Journeys from childhood to midlife: Risk, resilience, and

recovery. New York, NY: Cornell University Press.

WestEd. (2006). About the CHKS. Retrieved from http://chks.wested.org/about

WestEd. (2011). Resilience & youth development. Retrieved from http://chks.wested.org/

using_results/resilience

Whitlock, J. L. (2006). Youth perceptions of life at school: Contextual correlates of

school connectedness in adolescence. Applied Developmental Science, 10(1),

13-29.

Wigfield, A., & Eccles, J. (Eds.). (2002). Development of achievement motivation. San

Diego, CA: Academic Press.

Williams, T., Kirst, M., Haertel, E., Rosin, M., Perry, M., Webman, B., . . . Woodward,

K. M. (2010). Gaining ground in the middle grades: Why do some schools do

better. Mountain View, CA: EdSource.

Witherspoon, D., Schotland, M., Way, N., & Hughes, D. (2009). Connecting the dots:

How connectedness to multiple contexts influences the psychological and

academic adjustment of urban youth. Applied Developmental Science, 13(4),

199-216.

Wolin, S., & Wolin, S. (1993). The resilient self: How survivors of troubled families rise

above adversity. New York, NY: Villard Books.

146

Work, W. C., Cowen, E. L., Parker, G. W., & Wyman, P. A. (1990). Stress resilient

children in an urban setting. Journal of Primary Prevention, 11, 3-17.

Wyman, P., Cowen, E., Work, W., & Kerley, J. (1993). The role of children’s future

expectations in self-system functioning and adjustment to life stress. Development

and Psychopathology, 5, 649-661.

Zhao, Y. (2009). Catching up or leading the way: American education in the age of

globalization. Alexandria, VA: ASCD.

Zins, J. E., Elias, M. J., Greenberg, M. T., & Weissberg, R. P. (2000). Promoting social

and emotional competence in children. In K. Minke & G. Bear (Eds.), Preventing

school problems—Promoting school success: Strategies and programs that work

(pp. 71-100). Bethesda, MD: National Association of School Psychologists.

Zullig, K. J., Koopman, T. M., Patton, J. M., & Ubbes, V. A. (2010). School climate:

Historical review, instrument development, and school assessment. Journal of

Psychoeducational Assessment, 28, 139-152.

147

Appendix A

Select-Item Survey Questions

External - School Supports Scale (Caring Relationships and High Expectations

combined)

At my school, there is a teacher or some other adult ...

(1 = Not at All True, 2 = A Little True, 3 = Pretty Much True, 4 = Very Much True)

1. who really cares about me.

2. who tells me when I do a good job.

3. who notices when I’m not there.

4. who always wants me to do my best.

5. who listens to me when I have something to say.

6. who believes that I will be a success.

External - School Meaningful Participation Scale

At school…

(1 = Not at All True, 2 = A Little True, 3 = Pretty Much True, 4 = Very Much True)

7. I do interesting activities.

8. I help decide things like class activities or rules.

9. I do things that make a difference.

How true do you feel that these statements are about you personally?

(1 = Not at All True, 2 = A Little True, 3 = Pretty Much True, 4 = Very Much True)

Internal - Self-Efficacy Scale

10. I can work with someone who has different opinions than mine.

11. I can work out my problems.

148

12. I can do most things if I try.

13. There are many things that I do well.

Internal - Empathy Scale

14. I feel bad when someone gets their feelings hurt.

15. I try to understand what other people go through.

16. I try to understand how other people feel and think.

Internal - Problem-Solving Scale

17. When I need help, I find someone to talk with.

18. I try to work out problems by talking or writing about them.

Internal - Self-Awareness Scale

19. There is a purpose to my life.

20. I understand my moods and feelings.

21. I understand why I do what I do.

Healthy School Connectedness Scale (included in the CHKS)

How strongly do you agree or disagree with the following statements about your school?

(1 = Strongly Disagree, 2 = Disagree, 3 = Neither Disagree Nor Agree, 4 = Agree,

5 = Strongly Agree)

22. I feel close to people at this school.

23. I am happy to be at this school.

24. I feel like I am part of this school.

25. The teachers at this school treat students fairly.

26. I feel safe in my school.

149

Appendix B

2004 Regressions

Model Summary Without School Connectedness

R R Square F df1 df2 p .691 .477 103.2 10 1132 <.001

Coefficients Without School Connectedness

Source B SEM Standardized

Beta t P

(Constant) 516.054 38.668 13.346 <.001 supp -74.270 15.803 -.140 -4.700 <.001 part 168.028 15.370 .331 10.932 <.001 ps 9.873 12.102 .034 .816 .415 SE -9.939 12.074 -.043 -.823 .411

emp -3.492 12.772 -.014 -.273 .785 aware 37.054 11.863 .142 3.124 .002 pct_aa -2.169 .244 -.210 -8.893 <.001 pct_hi -1.449 .174 -.326 -8.307 <.001 meals -1.423 .146 -.317 -9.734 <.001

el .801 .282 .107 2.838 .005

Model Summary With School Connectedness

R R Square F df1 df2 P .708 .501 103.2 11 1131 .000

Coefficients With School Connectedness

Source B SEM Standardized

Beta t P

(Constant) 451.794 38.763 11.655 <.001 supp -136.363 17.569 -.257 -7.762 <.001 part 128.974 15.916 .254 8.104 <.001 ps 12.938 11.831 .044 1.094 .274 SE -4.915 11.816 -.021 -.416 .677

emp -10.013 12.509 -.040 -.800 .424 aware 32.388 11.607 .125 2.790 .005 pct_aa -1.683 .247 -.163 -6.809 <.001 pct_hi -1.505 .171 -.338 -8.825 <.001 meals -1.170 .147 -.261 -7.969 <.001

el .690 .276 .092 2.497 .013 conn 100.016 13.503 .258 7.407 <.001

150

Appendix C

2006 Regressions

Model Summary Without School Connectedness

R R Square F df1 df2 p .695 .483 91.2 10 976 <.001

Coefficients Without School Connectedness

Source B SEM Standardized

Beta t P

(Constant) 555.22 37.685 14.733 <.001 supp 3.050 16.319 .006 .187 .852 part 113.08 15.821 .233 7.148 <.001 ps 3.310 8.763 .016 .378 .706 SE 7.508 9.478 .044 .792 .428

emp 5.964 9.183 .035 .649 .516 aware -20.64 8.270 -.119 -2.495 .013 pct_aa -2.779 .250 -.277 -11.104 <.001 pct_hi -.749 .170 -.188 -4.415 <.001 meals -1.341 .141 -.331 -9.500 <.001

el -.165 .278 -.024 -.593 .553

Model Summary With School Connectedness

R R Square F df1 df2 P 0.730 .532 100.839 11.0 975 <.001

Coefficients With School Connectedness

Source B SEM Standardized Beta t P (Constant) 479.54 36.639 13.088 <.001

supp -90.794 18.090 -.180 -5.019 <.001 part 49.421 16.319 .102 3.028 .003 ps 1.580 8.342 .008 .189 .850 SE 14.136 9.045 .083 1.563 .118

emp 3.041 8.745 .018 .348 .728 aware -23.55 7.876 -.136 -2.990 .003 pct_aa -1.918 .253 -.191 -7.584 <.001 pct_hi -.843 .162 -.211 -5.217 <.001 meals -1.122 .136 -.277 -8.245 <.001

el .043 .265 .006 .162 .871 conn 142.64 14.096 .404 10.119 <.001

151

Appendix D

2008 Regressions

Model Summary Without School Connectedness

R R Square F df1 df2 p

.719 .518 88.5 10 825 <.001

Coefficients Without School Connectedness

Source B SEM Standardized

Beta t P

(Constant) 471.03 48.455 9.721 <.001 supp -2.662 20.904 -.004 -.127 .899 part 146.05 19.246 .267 7.589 <.001 ps 27.027 12.194 .136 2.216 .027 SE -35.977 11.791 -.224 -3.051 .002

emp 37.382 12.834 .226 2.913 .004 aware -13.03 10.448 -.078 -1.247 .213 pct_aa -2.769 .284 -.265 -9.742 <.001 pct_hi -1.065 .179 -.258 -5.938 <.001 meals -1.032 .162 -.256 -6.387 <.001

el .103 .302 .014 .342 .732

Model Summary With School Connectedness

R R Square F df1 df2 P 0.758 .575 101.472 11.0 824 <.001

Coefficients With School Connectedness

Source B SEM Standardized Beta t P (Constant) 416.34 45.786 9.093 <.001

supp -142.42 23.658 -.236 -6.020 <.001 part 84.076 18.996 .154 4.426 <.001 ps 25.373 11.450 .128 2.216 .027 SE -31.29 11.079 -.195 -2.824 .005

emp 29.29 12.073 .177 2.426 .015 aware -11.78 9.810 -.070 -1.201 .230 pct_aa -2.126 .274 -.203 -7.765 <.001 pct_hi -1.118 .168 -.271 -6.637 <.001 meals -.710 .155 -.176 -4.589 <.001

el .064 .284 .009 .225 .822 conn 172.98 16.351 .440 10.579 <.001