AN ASSESSMENT OF THE FACTORS THAT DRIVE PARENTAL CHOICE REGARDING OPEN ENROLLMENT AND INTRADISTRICT TRANSFERS Christopher Alan Morris B.S., California State University, Sacramento, 1995 M.A., National University, Sacramento, 2002 DISSERTATION Submitted in partial fulfillment of the requirements for the degree of DOCTOR OF EDUCATION in EDUCATIONAL LEADERSHIP at CALIFORNIA STATE UNIVERSITY, SACRAMENTO SPRING 2013 Copyright © 2013 Christopher Alan Morris All rights reserved ii AN ASSESSMENT OF THE FACTORS THAT DRIVE PARENTAL CHOICE REGARDING OPEN ENROLLMENT AND INTRADISTRICT TRANSFERS A Dissertation by Christopher Alan Morris Approved by Dissertation Committee: ________________________________________ Robert W. Wassmer, Ph.D. ________________________________________ Su Jin Jez, Ph.D. ________________________________________ Edward (Ted) Lascher, Jr., Ph.D. SPRING 2013 iii AN ASSESSMENT OF THE FACTORS THAT DRIVE PARENTAL CHOICE REGARDING OPEN ENROLLMENT AND INTRADISTRICT TRANSFERS Student: Christopher A. Morris I certify that this student has met the requirements for format contained in the University format manual, and that this dissertation is suitable for shelving in the library and credit is to be awarded for the dissertation. __________________________, Program Director Carlos Nevarez, Ph.D. iv ______________ Date DEDICATION This dissertation is dedicated to my wife, Lori, and daughters Olivia and Camille, and to the countless students that have inspired me with their dreams and aspirations. Without their support this would not have been possible. v ACKNOWLEDGEMENTS The unwavering support of Dr. Robert Wassmer was key in providing me the stamina to continue on this journey. Dr. Wassmer not only provided academic guidance, but also lifted my spirits when the path to completion was not always visible. From day one, Dr. Su Jin Jez forced me to produce only scholarship that I knew I could be proud of and to never doubt in my own ability to succeed. Dr. Jez was never reluctant to provide the coaching and feedback she knew I needed. Dr. Ted Lasher’s promise to make this a team effort is evident in the cohesiveness of the committee. His expertise in public policy and research helped me to turn an idea into policy recommendations with the potential to impact the lives of students and families across the state. Dr. JoLynn Britt has been an enthusiastic supporter of my work. Dr. Britt helped turn the daunting task of multiple types of analysis into an endeavor that allowed me to grow professionally and academically. As this journey comes to an end, I would like to express my sincere thanks and appreciation to my extended family in cohort four. This amazing group of educators and leaders will have an effect on public education for years to come. vi CURRICULUM VITAE Education M.A. Educational Administration, National University, Sacramento B.S. Criminal Justice, California State University, Sacramento Professional Employment Vice Principal/Director, Vallejo City Unified School District, Vallejo, California Principal, San Juan Unified School District, Carmichael, California Principal, Plumas Lake Elementary School District, Plumas Lake, California Vice Principal, Elk Grove Unified School District, Elk Grove, California Fields of Study Open enrollment Education policy Adolescent psychological, social, and emotional development vii Abstract of AN ASSESSMENT OF THE FACTORS THAT DRIVE PARENTAL CHOICE REGARDING OPEN ENROLLMENT AND INTRADISTRICT TRANSFERS by Christopher A. Morris With increased school choice options for parents, open enrollment in public schools has seen significant growth. As parents seek greater access to educational options for their children, urban schools and schools in high poverty areas have seen enrollments plummet. This has led to a decrease in the educational opportunities for disadvantaged students that remain at their home schools. Research on school choice has focused primarily on charter schools and private schools, with minimal attention given to the factors that lead parents to choose a regular public school other than their home school within the same school district. With nearly 20% of all California students in grades K-12 exercising some form of school choice, this study will more closely examine the phenomenon of families choosing to enroll their children in regular public schools other than their home school. For this study, quantitative methods are used. Within the frameworks of rational choice theory and social cognitive theory, this study answers the following questions: 1) to what extent do schools have the capacity to affect the changes necessary to influence enrollment patterns of their schools, and 2) to what degree do current school policies influence open-enrollment and school choice decisions of parents? viii The participating school district for this study is a large comprehensive pre K–12 school district in northern California. The data for this study included demographic and enrollment data for the entire study body. The second source of data was derived from parent surveys of participants and non-participants in open enrollment. The results of this study suggest that the greatest indicator of participation in open enrollment is if the neighborhood school has a lower API than the district’s average. Additionally, students that identify as “other Asian” and not specifically with one of the groups identified, are White and speak a language other than English, or are female, also have an increased likelihood of participation in open enrollment. Parents also indicated that the availability of desirable programs was a significant factor in open enrolment participation. Keywords: open enrollment, school choice, transfer students, charter schools, intradistrict, interdistrict, student achievement, secondary education, urban education ix TABLE OF CONTENTS Page Dedication ............................................................................................................................v Acknowledgements ............................................................................................................ vi Curriculum Vitae .............................................................................................................. vii List of Tables ................................................................................................................... xiii List of Figures ....................................................................................................................xv Chapter 1. SCHOOL CHOICE AND STUDENT OUTCOMES .....................................................1 Introduction ..............................................................................................................1 Open Enrollment ......................................................................................................5 Impact on the Poor, People of Color, and Urban Communities ................10 Nature of Study & Research Questions .................................................................13 Theoretical Frameworks ........................................................................................15 Rational Choice Theory .............................................................................15 Social Cognitive Theory ............................................................................17 Significance of the Study .......................................................................................20 Remainder of the Study .........................................................................................21 2. REVIEW OF RELATED LITERATURE ....................................................................23 Social and Cultural Aspects of School Choice ......................................................24 Charter Schools ......................................................................................................38 Effects of Choice on Student Learning ..................................................................41 x Impact of Choice on Non-Choice Schools.............................................................56 Continued Debate on School Choice .....................................................................57 Conclusion .............................................................................................................61 3. METHODOLOGY .......................................................................................................64 Introduction ............................................................................................................64 Research Questions ................................................................................................64 Population and Setting ...........................................................................................65 Data ........................................................................................................................67 Quantitative Methods Used in this Study ..............................................................71 Logistic Regression Method ......................................................................72 Survey-Based Method ................................................................................74 Sampling ....................................................................................................77 Role of the Researcher ...........................................................................................79 Protection of Participants .......................................................................................79 4. DATA ANALYSIS AND FINDINGS .........................................................................81 Introduction ............................................................................................................81 School District Data ...............................................................................................82 Descriptive Statistics ..................................................................................82 Logistic Regression ....................................................................................87 Analysis of Survey Results ....................................................................................91 Descriptive Statistics ..................................................................................91 Test For Equality of Means........................................................................96 xi Summary ..............................................................................................................101 5. CONCLUSIONS AND RECOMMENDATIONS .....................................................102 Introduction ..........................................................................................................102 Overview ..............................................................................................................103 Research Questions ..............................................................................................103 Summary of Findings ...........................................................................................104 Discussion ............................................................................................................114 Policy Implications ..............................................................................................117 Recommendations ................................................................................................118 Future Research ...................................................................................................125 Appendix A. Probability in Finding a School Attribute Important ................................127 Appendix B. Survey to Parents Participating in Open Enrollment .................................128 Appendix C. Survey to Parents Not Participating in Open Enrollment ..........................129 Appendix D. GUSD Open Enrollment Application .......................................................130 References ........................................................................................................................131 xii LIST OF TABLES Page 2.1. School Attribute Rankings of 1,582 Parents in New York City Suburbs ..................38 2.2. Black Over-Representation in Charter Schools; Top 15 MSAs 2007-2008 ..............39 2.3. The Race and Gender of New York City Charter School Applicants and Students in Traditional Public Schools from 2000 to 2006 .....................................................40 3.1. Primary Languages of Students Within GUSD ..........................................................68 3.2. Ethnic and Racial Categories of Students Within GUSD ...........................................69 3.3. English Proficiency Designations of Students Within GUSD ....................................69 3.4. Descriptions of Variables in Survey ..........................................................................75 4.1. Comparisons of Means of Students Characteristics Between Participants and Non-Participants ........................................................................................................84 4.2. Logistic Regression Results .......................................................................................88 4.3. Statistically Significant Effects on Participation in Open Enrollment.......................90 4.4. Fit of the Logistic Regression Prediction...................................................................91 4.5. Descriptive Statistics of Survey Responses (Questions 1-5) .....................................93 4.6. Descriptive Statistics of Survey Responses (Questions 6-14) ...................................94 4.7. Descriptive Statistics of Survey Responses (Questions 15-18) .................................95 4.8. Comparisons of Means of Survey Responses (Questions 1-14) ................................99 4.9. Comparisons of Means of Survey Responses (Questions 15-18) ............................100 5.1. Programs that Influence Parents’ Decision to Stay at Their Neighborhood School .....................................................................................................................111 xiii 5.2. Programs that Influence Parents’ Decision to not Enroll Their Child at the Neighborhood School .............................................................................................112 xiv LIST OF FIGURES Page 1.1. K-12 Enrollment Over Time in California...................................................................6 1.2. Percentages of Students in School Type by Race / Ethnicity ......................................7 1.3. Numbers of Charter Schools in California from 1999 to 2008 ....................................9 1.4. Percentage of Students Enrolled by School Type from 1993 to 2007 .......................10 1.5. Math Scores Among 4th Grade Students in the U.S. from 2003 to 2011 ...................12 1.6. Math and Reading Performance of U.S. Class of 2011 by Race/Ethnicity ...............13 1.7. Factors Central to Social Cognitive Theory...............................................................19 2.1. Quality and Impact Ratings for Studies of Achievement in Magnet or Interdistrict Choice Schools..........................................................................................................45 2.2. Percentage of Schools that Meet Half or More of Their Final Goals ........................50 2.3. Comparisons of the 4th Grade Students Performing at or Above Connecticut Goal on CMT Writing Test .......................................................................................54 2.4. Comparisons of the 6th Grade Students Performing at or Above Connecticut Goal on CMT Writing Test .......................................................................................54 2.5. Comparisons of the 8th Grade Students Performing at or Above Connecticut Goal on CMT Writing Test .......................................................................................54 3.1. District Racial and Ethnic Distribution for 2010 – 2011 School Year ......................65 3.2. District Students Percent at or Above Proficient – Math 2008 – 2012 ......................66 3.3. District Students Percent at or Above Proficient – English Language Arts 2008 – 2012...........................................................................................................................66 xv 3.4. Enrollment Patterns at Study High Schools ...............................................................77 4.1. Comparisons of API Score of O.E. Participants and Non-Participants .....................85 4.2. Comparisons of Home Language of O.E. Participants and Non-Participants ...........85 4.3. Comparisons of Socioeconomic Status of O.E. Participants and Non-Participants ..85 4.4. Comparisons of Female of O.E. Participants and Non-Participants ..........................85 4.5. Comparisons of White, Non-English Home Language O.E. Participants and Non-Participants .......................................................................................................86 4.6. Comparisons of Hispanic or Latino O.E. Participants and Non-Participants ...........86 4.7. Comparisons of South-East Asian O.E. Participants and Non-Participants ..............86 4.8. Comparisons of Other Asian O.E. Participants and Non-Participants.......................86 5.1. Characteristics Likely to Increase or Decrease Participation in Open Enrollment ..106 5.2. Graphical Representation of Numerically Significant Factors that Influence Open Enrollment Decisions – From Logistic Regression ................................................107 5.3. Mean Scores of Statistically Significant Parent Perceptions Regarding Academic Programs .................................................................................................................109 5.4. Conceptual Framework of Educational Goal Attainment ........................................124 xvi 1 Chapter 1 School Choice and Student Outcomes Introduction Open enrollment in public K-12 schools has gained significant momentum in the United States over the past ten years. According to the National Household Education Surveys Program (NHES, 2010), 80% of students were attending their school of residence in 1993, whereas that number had declined to 69% in 2007. The study indicates that while the number of students enrolling in private schools and home-schools has increased, the largest percentage of families seeking alternatives did so by using some form of school choice (NHES, 2010). It is important to note that interdistrict open enrollment serves more students than any other type of school choice program (Center for Education Reform, 2009). In a recent Gallop Poll of Americans’ confidence in public schools, 29% responded they had a great deal or quite a lot of confidence. That is a new low since that question has been asked since 1973. This represents a disparity, however, in how Americans perceive their neighborhood schools. Further, 37% of the Gallup Poll respondents gave their child’s school a grade of A, whereas only 14% gave public schools as a whole a grade of A (Gallup, 2011). This shows that the public’s confidence in the once highly regarded public school system has eroded. Much of the erosion of support for the neighborhood school model is a result of government policies intended to help the increasingly diverse population of students that are potentially harmed by open enrollment policies. 2 Although school choice is growing across the nation, California and the United States’ mediocre performance worldwide in mathematics and reading has most people agreeing that more public education reforms are needed. The Organization for Economic Cooperation and Development administers the Program for International Student Assessment (PISA) to representative samples of 15-year-old students from 65 school systems around the world. Based on the most recent data from the class of 2011, U.S. students ranked 32nd in math among those countries that participated. U.S. students fared better in reading, with only 10 other countries outperforming the United States by a numerically significant amount (Peterson, Woessmann, Hanushek, & Lastra-Anadon, 2011). The same study showed that students in California ranked 40th out of all the states and District of Columbia. Viewpoints about reform include greater accountability, increased educational requirements for teachers, increased local control, and school choice options. The larger question is to what end are these reforms intended to produce? There are the benefits for each individual student and then there are the public benefits and improved social outcomes. What is known through the research is that students who receive better schooling are able to transfer their skills into higher income and better occupations, they benefit from greater self-understanding, have more positive personal contacts, and enjoy a healthier lifestyle (Levin, 2012; Culter & Lleras-Muney, 2008). The societal benefits, on the other hand, are the health, legal, political, social, and economic benefits that come from an educated society. What is also known is that higher achieving schools provide 3 greater access to a more rigorous curriculum, guidance in course-taking patterns, and access to extracurricular activities (Lauen, 2007). Culter & Lleras-Muney (2008) further explore the link between health, education and wealth. With the knowledge that increased education leads to greater wealth, their study of adults over the age of 25 found that as wealth and education increased, there was a decrease in the use or abuse of alcohol, tobacco, and illegal drugs. Additionally, with increased education, adults over 25 had reported increased exercise, household safety, and access to preventive health care including flu shots, mammograms, and colon cancer screening. Further, individuals with high social hierarchy (education, wealth, employment) enjoy better health than those below and have lower rates of mortality and morbidity (Adler, et al., 1994). When examining the positive benefits of the investment one makes in education, researchers have agreed that the effect of an educated community can be seen in better public health, lower population growth rates, greater volunteering and financial giving, political stability, lower homicide rates, lower property crime rates, and greater participation in democracy. The effects are not always positive however. As wealth increases as a direct result of increased education, individual spending and economic choices change. Prior to open-enrollment policies, parents have been exercising choice through housing. With earnings rising by approximately 10% for each additional year of post secondary education, (Krueger & Lindahl, 2000), urban sprawl and an exodus to the suburbs have been a result of families looking to escape urban schools. Negative externalities as a result of this include increased air pollution and driving times, traffic congestion, need for 4 more roads and infrastructure, a depletion of open spaces, an increase in the concentration of the poor in the cities, and a further decline in the quality of the public schools in these centralized communities (Wassmer 2006 and McMahon 2004). With nearly 20% of all California students in grades K-12 exercising some form of school choice, this study will more closely examine the phenomenon of families choosing to enroll their children in regular public schools other than their neighborhood school. It will not examine the factors related to parents choosing to enroll their children in private schools or home school programs. Although research exists on why parents choose private schools over public schools, far less is known as to why parents choose to enroll their child in a regular public school other than their own within their school district (Dillon 2008; Grady 2010; Tenbusch 1993). This dissertation study identifies the factors that influence a parent’s decision to enroll their child in a public school, within the same school district, other than their neighborhood school. Additionally, this study provides valuable information to education policy-makers to inform open enrollment and school improvement policies. Lastly, school leaders can use the results of this study to determine the extent to which their schools have the capacity to affect the changes necessary to influence enrollment patterns of their schools. This first chapter offers the necessary background for this dissertation. It does this through the following. It first provides historical background on open enrollment in California by examining current laws and analyzing enrollment trends of the nearly seven million California K-12 students. Second, I will provide a detailed description of the problem and rationale justifying further study. Next, there is a thorough discussion of the 5 theoretical frameworks used to analyze the phenomenon of intradistrict transfers in California. Through the lenses of rational choice theory and social cognitive theory, I examine the complexities of choice and human decision-making. Chapter one concludes with a description of each of the remaining chapters. Open Enrollment In recent decades, more and more of California’s families have exercised some form of school choice for their children. It is important to note that in school districts with open enrollment policies, it could be argued that even those families that stay at their neighborhood schools are in essence exercising the choice to stay. For the purposes of this study, choice schools, or families exercising choice refers only to those students and families that have chosen to leave their neighborhood school. The are nearly seven million K-12 students in California enrolled in various school types. Figure 1.1 shows the distribution of California K-12 students in private, charter, and regular public schools. What is noteworthy about this chart is the rising bar representing charter school enrollment. Though charter school enrollment is not what is specifically studied here, it does offer an indicator of Californians’ dissatisfaction with their neighborhood public schools. In 1998 – 1999 charter schools represented 1% of California K-12 enrollment. By 2008 – 2009, this had grown to 4.2%. In actual numbers there was an increase of 217,000 students enrolled in public charter schools. 6 Figure 1.1. K-12 Enrollment Over Time in California Increasing numbers of school districts are providing families with the option to exercise a form of school choice through their open-enrollment policy. More specifically, many school districts allow for voluntary transfers to any age appropriate elementary, middle, or high school within their district. The term used to describe these transfers is intradistrict, meaning within. Conversely, interdistrict transfers describe students attending schools in districts outside of their home school district. According to the Policy Analysis for California Education PACE (1999), nearly one fifth of all children nationwide, about seven million, no longer attend their neighborhood public school, with nearly ten percent attending private schools (Fuller, Burr, Huerta, Puryear, & Wexler, 1999). While open enrollment advocates sometimes argue that these policies will most benefit students from lower socioeconomic backgrounds, it is white, affluent and working-class families that exercise choice options with the greatest frequency. For 7 example, 22% of the students enrolled in a public school of choice were poor, 21% were near-poor, and 57% percent were non-poor. Figure 1.2 goes on to show that while the percentage of Black and Hispanic students enrolled in schools of choice is greater than their proportion enrolled in neighborhood schools, it is still White students by a ratio of more than 2:1 that make up students enrolled in schools other that their neighborhood schools (National Household Education Surveys Program, 2010). Figure 1.2. Percentages of Students in School Type by Race/Ethnicity California’s policies related to open enrollment center around providing choice to families with children in failing schools. In 2010, California passed the Open Enrollment Act (SBX54), more commonly referred to as the Romero Bill. The intent of the act is to provide an option for students enrolled in one of the 1,000 “low-achieving” schools in the state to enroll in a different school with a higher Academic Performance Index (API) than the pupil’s school of residence (California Department of Education, 2011). The purpose of the Open Enrollment Act is to “improve student achievement and enhance parental choice in education by providing additional options to pupils to enroll in public schools throughout the state without regard to the residence of their parents” 8 (http://www.cde.ca.gov/sp/eo/op/faq.asp). The list of 1,000 schools consists of 687 elementary schools, 165 middle schools, and 148 high schools. Excluded from the list are court schools, community day schools, schools with fewer than 100 valid scores, and charter schools. Additionally, no LEA may have more than 10% of their total schools on the list. California’s open enrollment law goes a step further than the mandates of the No Child Left Behind Act of 2001. NCLB states that only students attending Title I schools that have been identified as in need of improvement are eligible for transfer. California does not have the Title I requirement (U.S. Department of Education, 2007). Parental choices are private schools, charter schools, vouchers, magnet schools, homeschooling, dual/concurrent enrollment, and more commonly, intradistrict and interdistrict transfers. Presently, open-enrollment policies exist in 46 states with many dating back to the 1960s following the Brown v. Board of Education decision outlawing de jure, or lawful, segregation. In practice, however, a segregated system is often left in place when white students use these policies to leave desegregated schools in favor of allwhite schools (Dillon, 2008). While charter schools represent the smallest number in terms of public school options, their growth over the past decades illustrates the growing demand for more school choice options. In 1992 California was the second state in the nation to authorize charter schools. Figure 1.3 illustrates the growth in charter schools over the past decade. States throughout the nation have experienced similar growth in the number of charter schools and charter school enrollment. A reference back to Figure 1.1 shows that the growth in charter schools has occurred at a similar pace as the increase in 9 total charter school enrollment. As a result, charter schools are educating a growing percentage of students in California schools. Figure 1.3. Numbers of Charter Schools in California from 1999 to 2008 It is widely understood that charter schools and regular public schools have different characteristics. Therefore, when examining choice, the two do not compare well with each other. California Education Code section 35160.5(b) specifies that intradistrict open enrollment is locally determined, and that each local educational agency (LEA) must adopt rules and regulations establishing their open enrollment policy. Commonly, schools allow for transfers-in providing there is space available. The California School Boards Association, however, recommends each district maintain a 5% reserve of the school’s capacity to accommodate students of residence. Figure 1.4 further illustrates the growth in school choice from 1993 – 2007. 10 Figure 1.4. Percentages of Students Enrolled by School Type from 1993 to 2007 The minimal research available examining the perceptions of parents utilizing their districts’ choice options combined with the 5% growth in enrollment for public schools of choice, creates the opportunity for the results of this study to greatly inform school policies and have a positive impact on the educational outcomes of students from culturally and economically diverse backgrounds that choose to remain in their neighborhood schools. Impact on the poor, people of color, and urban communities. The number of factors and challenges often experienced by low income and minority families in attaining a quality education for the children, and California’s increasing diversity, underscore the urgency in preparing our young people for future success. With Black and Latino students comprising 57% of the student population, Asian and Filipino students an additional 12.5%, and the increasing urbanization in California, the effort to address the educational achievement of students in our most needy schools takes on greater urgency. Given this performance, and the growth of open- 11 enrollment policies, policy makers must continue to examine ways to improve student performance and educational outcomes for poor and minority students. Increased open-enrollment options in California, and the exercise of this school choice has led to declining enrollment in many public neighborhood schools. This declining enrollment has especially occurred in the state’s urban public schools that disproportionately serve low income and people of color. As a result, the potential exists for a greater loss of programs in these types of schools leading to a further decrease in the educational opportunities for the disadvantaged students that remain at their neighborhood schools. With increased deregulation of school assignment policies, home address is no longer the sole determining factor in school assignment. Families with greater financial means are able to access choice options more readily than families in poverty (Lauen, 2007). The challenge of concentrated poverty and shrinking resources for public schools further compels policy makers and educators to examine current policies in order to better meet the needs of students in these communities. When low-income students are concentrated and separated from their middle-income and middle-class peers, there tends to be less academic support. Moreover, middle-class schools are twenty-two times more likely to be high performing than schools in high-poverty areas (Kahlenberg, 2006). Scores from the 2011 National Assessment of Educational Progress shows that while students from all economic groups showed academic gains, there still exists a gap between students of poverty and middle-class students. The chart in Figure 1.5 illustrates this gap in math achievement. Further, the chart also shows that with the growth of 12 school choice options, the achievement gap between those students in and out of poverty remains consistent. It suggests that poverty is the greatest predictor of low academic achievement regardless of school choice. Figure 1.5. Math Scores Among 4th Grade Students in the U.S. from 2003 to 2011 As recent as 2011, 42% of white students were identified as proficient in math while only 11% of African American students, 15% of Hispanic students, and 16% of Native American students were identified and proficient. Figure 1.6 shows the math and reading performance of the United States’ class of 2011 (Peterson, Woessmann, Hanushek, & Lastra-Anadon, 2011). This data further illustrates the gap that exists between white students and students of color. As parents of minority students see their children fall further behind their peers, it creates an even greater sense of urgency in seeking educational options that they believe give their children increased access to quality education. 13 Figure 1.6. Math and Reading Performance of U.S. Class of 2011 by Race/Ethnicity Nature of Study & Research Questions This study involves both qualitative and quantitative strategies. The quantitative analysis involved using logistic regression and means comparison. The qualitative strategy analyzed the survey responses from participants and non-participants in open enrollment. I used these strategies to explore the demographic characteristics of those who choose to participate, and not participate in open enrollment, and to study the first hand perceptions of parents at the center of the phenomenon of enrolling their children in public schools other than their neighborhood schools. The research questions are answered using both research methods. In order to understand the perceptions of parents choosing to enroll their child in a school other than their home school, this study specifically perform these research tasks: 1. Using logistic regression analysis of a dependent variable measuring whether a student left his or her neighborhood school for another school in the same district, 14 I uncover what variables (available in standard data files) exert a significant impact on this choice. This analysis allows me to predict which variables are related to parent decision-making regarding school choice. 2. I also analyze results from a survey or parents. This information allows me to probe more deeply into the motivations for school choices, thereby helping me to make policy recommendations. The purpose of this study is to uncover the factors leading to the phenomenon of parents moving their children from neighborhood schools and use that information to inform open enrollment and school improvement policies. Additionally, this study seeks to uncover the extent to which schools have the capacity to change or influence enrollment patterns of their schools. I am to answer the question: To what degree do current school policies influence open enrollment and school choice decisions of parents? The groups of parents to be analyzed in this study are those parents who have chosen to enroll their child in a regular public school other than their assigned school of attendance. Researchers have underscored the need for policy makers and school officials to thoroughly understand the dynamics of school choice by making the argument that successful policies cannot be implemented without a thorough theoretical understanding of the traditional school system and the implications to families that choose to go outside that system (Koedel, Betts, Rice, & Zau, 2009). 15 Theoretical Frameworks The complexities of human choice and the ways in which individuals express their preferences in schools for their children may only be partially quantifiable. However, through the application of theories used in the fields of economics, and psychology, I posit how parents’ perceptions influence those choices. These theoretical approaches are described next. Rational Choice Theory. Rational Choice Theory (RCT), originally developed by economists, is useful in understanding human decision-making based upon their maximizing self-interest. In a broad definition of rationality, it implies a degree of sensibleness and logic. Rational Choice Theory, however, uses a more narrow definition of "rationality" to mean that an individual acts to maximize their personal advantage (Abell, 1990). It takes into account motivation, goal attainment, and personal advantage. The theory of Rational Choice will help to explain and understand this specific human behavior and the phenomenon of parents choosing to enroll their children in schools other than their neighborhood school. Further, Rational Choice Theory attempts to examine how those individual private choices produce outcomes that may have social consequences. Those consequences will be described as the negative externalities occurring as a result of their choice. Under Rational Choice Theory, it is expected that parents do not take into account these negative externalities when making a choice to pull a child out of a neighborhood school and move them to another school. Only when parents use this in deciding to leave or stay 16 at their neighborhood school cold they then assess the benefits and costs of making this choice. While Rational Choice Theory is the basis of economics, in recent decades researchers have expanded its use to sociology, political science, and anthropology (Green S., 2002). At its most basic elements, Rational Choice Theory attempts to analyze individual or group behavior and examine how those individual choices produce outcomes. The simplest model examines the relationship between buyers and sellers. This study, however, will supplant the buyers with parents of school-age children, and sellers with schools and school districts. Additionally, the readers must acknowledge the role of parents as consumers (Frankenberg, Siegel-Hawley, & Wang, 2010). Denzin (1990) cites the early work of many authors when he argues that the following assumptions are at the center of Rational Choice Theory. First, humans are egotistical, hedonistic, asocial, rational, and purposive in their actions. Secondly, when choosing the most rewarding course of action, emotions and sentiments are factors. Third, choices are made based on principle, information, history, an assessment of the consequences, and weighing the chances of achieving the desired effect. This means that in order for a rational choice to be made, one must know the desirable and undesirable consequences of a choice, and must know how the choice may affect other plans. Most importantly, this choice cannot violate any of their held principles. The choice is also made within the context of the individual or family, with little to no consideration of a broader social impact. 17 The fundamental question is how do parents of school age children as consumers of the public education services offered to their children reach decisions on where to attend? While some decisions are based on custom or habit, rational choice premises that choices are made that best help both buyers achieve their personal objectives. Due to a lack of full information on the benefits and costs involved in different choices, and the expense to the family of gathering this information, decisions made about schooling may revert to relying on custom or habit. For example, a parent may say, “the school is where all my other children went, so why not send my youngest there,” Or “this is the school that my children’s friends are attending, I will also send her there.” Much of the purpose of this study is to help uncover the factors parents consider when selecting a school for their child. There will also be consideration of the limitations of rational choice as a guide to public policy. Social Cognitive Theory. Social Cognitive Theory is an additional framework that will be applied to this study. In contrast to RCT, which fails to consider human group life, emotions, or human experience (Denzin, 1990), Social Cognitive Theory suggests that people model their behavior after others they identify with, and deals with the cognitive and emotional aspects of behavior. It is often used to explain behavior patterns and evaluate behavioral change based on environment, people, and behaviors. As Social Cognitive Theory relates to this study, it is hypothesized that parents’ perceived social status changes by enrolling their child in another school. It follows that 18 this perceived change in social status and self-identification with a social group will transfer into a positive effect on student learning outcomes and increased child’s selfefficacy. Although this study does not present the empirical evidence in support of this claim, it is widely held that exposure to economic resources, social prestige, and social power have a positive influence on learning experiences (Thompson & Dahling, 2012). The core assumption regarding social cognitive theory is that environment, people, and behavior are three factors that are constantly influencing each other. Further, according to Bandura (1997), a person’s actions are not only reinforced by the actions of others, but also by their environment. Additionally, of the concepts of SCT identified by Baranowski et al. (1997), the following are most use useful as they relate to this study: Environment – provides opportunities for social support Self-control – provides opportunities for self-monitoring, goal setting, problem solving, and self-reward Observational learning – behavioral acquisition that occurs by watching the actions and outcomes of others Reinforcements – promotes self-initiated rewards and incentives (Glanz et al, 2002, p.169). The conceptual model in Figure 1.7 further illustrates the interchange between the factors central to SCT (Pajares, 2002). 19 Figure 1.7. Factors Central to Social Cognitive Theory Similarly, there is a degree of self-identification and self-stereotyping that occurs when children are made aware of social grouping (Johnson & Shapiro, 2003). Bennett and Sani (2008) hypothesize that when individuals categorize themselves as members of a social category, a process of depersonalization takes place. More specifically, depersonalization causes members of the same group to begin to develop a sense of similarity and will accentuate those traits that distinguish themselves from other groups. Moreover, new members to the group will begin to adopt those group-specific traits. While there will be little discussion of the interplay between competition and market forces in education, it is important to note that a key assumption of proponents of school choice is that competition will improve the level of schooling for all. Further research may explore the hypothesis that cooperation among schools and school districts can have an even greater positive effect on the educational outcomes of students. 20 Significance of the Study This study is significant because there are a growing number of socioeconomically disadvantaged students unable to access current school choice options. This leaves them stuck in schools where the threat of exit is improbable. Where school choice is exercised, there are also a large number of impoverished parents who remain at a neighborhood school because they do not have the resources to exercise choice (de Souza Briggs, 1997; Dee, 1998; Public Policy Forum, 2002). Two imperatives for California school leaders have emerged. First, for school leaders in urban schools and schools whose geographic isolation does not allow for parents to exercise school choice, They must have adequate information in order to improve the educational outcomes for all of their students. The second imperative is for school leaders understand this phenomenon better so they can improve the characteristics of the schools that drive parents away. Since interdistrict transfers account for the majority of school choice options, less data is available on the impacts of choice when students stay within their districts (intradistrict transfer). In these cases, the imperative is not to help people to exercise choice; rather it is to create school environments that will make people want to stay. During this time when the need for high-skilled workers has grown and high rates of unemployment have persisted, educators and policy makers must use every tool to ensure that America’s young people have the skills and abilities necessary to succeed in the global economy of the 21st century. 21 Remainder of the Study This dissertation contains five remaining chapters. This chapter introduced the study, contained a description of the problem, described the nature of the study, summarized the theoretical frameworks, provided operational definitions, outlined assumptions and limitations, and the significance of the study. Briefly restated, the problem is that as open enrollment and school choice continue to grow, students from economically and culturally diverse backgrounds face even greater challenges when they remain in “hollowed out” public schools. While students in the United States and California continue to show gains on standardized tests, the gap between poor and nonpoor students, and between white students and students of color remains wide. Through the theoretical lenses of Rational Choice Theory and Social Cognitive Theory, I will examine the perceptions of parents that choose open-enrollment options for their child. The study will consist of both qualitative and quantitative research questions. Chapter 2 provides a review of relevant literature on open enrollment, social and cultural aspects of school choice, and achievement effect of school choice. The psychological and social influences of school choice are also examined with an emphasis on the effect of choice on those that remain at their neighborhood schools. Chapter 2 also includes a brief discussion of the Cultural Ecology Theory and the characteristics that have accompanied minority schooling in America. Chapter 3 details the methodology used in this study. In Chapter 3, sampling procedures, data collection and analysis, and issues related to validity and reliability are presented. 22 Chapter 4 presents, explains, and interprets the data. I begin by presenting the analysis of the logistic regression from the data provided by the district. The logistic regression identifies those variables that exert a statistically significant impact on the choice to participate in open enrollment to not participate. Next is the correlation and mean comparison showing which factors (available through a self-administered survey of parents in this school district) are significant in relation to why families choose to participate in open enrollment and leave their neighborhood school, or choose to remain at their neighborhood school and not participate in open enrollment. Lastly, through the analysis of written responses, I look for any patterns that may indicate to what degree do current school policies influence open enrollment and school choice decisions. Chapter 5 summarizes the major findings and outlines areas for further research. I provide further explanation of those variables I found statistically significant in relation to school choice decisions. In Chapter 5 I revisit each of the research questions and specifically describe how each question was addressed. Chapter 5 concludes by providing recommendations for policy makers and K through 12th grade school leaders. 23 Chapter 2 Review of Related Literature The vast majority of research on school choice examines enrollment patterns, student achievement, specific demographics of families opting for choice, or seeks to evaluate charter or voucher programs. There is little research on the psychosocial influences on parents when making decisions about schools for their child. This study closely examines the attitudes, perceptions, and influencers of parents who choose to enroll their child in a school other than their home school. The questions central to this study are (1) what variables, if any, significantly influence parents’ choice and selection of school for their child during open enrollment, (2) to what extent do schools have the capacity to affect the necessary changes to influence enrollment patterns, (3) and to what degree do current school policies influence open enrollment and school choice decisions? The following subsections categorize the literature reviewed in this chapter: (a) social and cultural aspects of school choice, and (b) achievement effects of school choice. Much of the data presented comes from California schools, however, the issues raised and the questions posed are representative of large urban school districts across the entire country. I specifically chose to examine the following themes and subjects for this literature review: (a) open-enrollment, (b) school choice, (c) school competition, (d) school reform, (e) the economic impact of school choice and open enrollment, (f) urban schooling, and (g) charter schools. The vast majority of the literature uncovered statistics 24 and data resulting from quantitative research. There is a significant body of knowledge related to student outcomes and achievement in public, charter, and private schools. Additionally, many studies have been conducted to evaluate the effectiveness of school choice models in the areas of racial integration, student achievement, access to programs, and efficiency. This study extends the literature by examining the psychological and social perceptions that influence parent choice decisions. Additionally, this study will explore the relationship between rationality and social influences when making school choice decisions. Articles and studies published in peer-reviewed journals account for the majority of the literature reviewed in this chapter. However, additional data was gathered from state departments of education, the U.S. Department of Education, university studies, and from papers published by seminal researchers on school choice and economic theory. Social and Cultural Aspects of School Choice Prompting many to advocate for school choice is the disproportionate number of non-white children at risk of dropping out of high school, unemployment, imprisonment, and poverty. This resulted in greater efforts to try to improve the educational outcomes of this group. One such method, though controversial, is through increased school choice options. In general, choice refers to decision-making when faced with two or more possibilities. As it relates to this study, choice is used to describe the right of parents to be able to choose where to send their children to school. Parents exercise school choice through enrollment in charter schools and private schools, through participation in school 25 voucher programs, and through school district open-enrollment programs. In accordance with a 1993 state law, California public school districts have created intra- and interdistrict public school choice policies, whereby a student may choose to attend a participating school outside the student's neighborhood if space permits. Under the federal No Child Left Behind Act (NCLB), school districts must allow students to transfer out of consistently low-performing or persistently dangerous schools, as defined by the state (EdSource, 2012). The controversy on the use of school choice is primarily fought on two fronts. First, it is argued without adequate social and financial resources, families in disadvantaged neighborhoods are less likely to seek, apply for, and travel to schools of choice (Lauen, 2007). Secondly, school choice has the potential to cream the best students from schools and leave those hollowed-out and filled with the weakest students and fewer economic resources. Segregation in metropolitan areas concentrates poverty, intensifies segregation in schools, and isolates the most disadvantaged children limiting their access to educational and economic opportunity (Orfield & Wallace, 2007). Often, de facto segregation occurs as a result of residential patterns of Black and Latino families and a lack of affordable housing within the boundaries of the most effective schools. A rarely discussed consequence of public school integration efforts is the toll it takes on communities and families. With students often traveling more than an hour to school, it not only creates physical distance, but social distance to the community. The disconnect that occurs when children do not go to school with their neighbors, or are not able to participate in the 26 kinds of extracurricular activities that happen after school, has a disproportionately negative effect on disadvantaged African American students and families. White students and students of higher socio-economic status that live within the boundaries of the choice schools do not experience this community disconnect (Smrekar & Goldring, 1999). An ideal model of integration would include integrated neighborhoods where Blacks and whites live next to each other and there is a real opportunity for discourse, communication, and understanding. Sending kids off to integrated schools has many convinced that any gains are temporary and insufficient to bring about any lasting change. From the point of view of teachers, by eliminating the schools of already disadvantaged Black children and bussing them to other communities, they risk loss of their culture, their history, and their heritage (Morris, 1997). When asking parents, however, to consider the personal costs of bussing, they cited the benefits of diversity, tolerance, and integration as being worth it (Smrekar & Goldring, 1999). Notably, what parents described as good schools tended to be in predominately white suburbs (Johnson & Shapiro, 2003). For many, schools, housing, and race are intertwined in such a way that a discussion of one leads to a discussion of the others. When considering school choice, parents with financial means continue to exercise school choice by residence to a degree that families in poverty could ever hope to do. When Johnson & Shapiro (2003) asked white parents to explain how they chose their neighborhood, most parents explained right away that it was because of the school or school district. Their interviews revealed that when asked to explain what “good neighborhoods” and “good schools” meant to them, parents’ explanations usually 27 included a discussion of race. When asked if race played a role in any of the decisions she had made, a white parent replied, “I have to be honest and…I’m probably wrong for saying it, but truthfully, it’s in the back of my mind, yes.” She goes on to say, “if there was a nice Black family…and they bought the house next door to us and had the same values and the same desires and goals that we had…I wouldn’t be afraid to have my children carpool or sit by them. I guess I am a racist deep down inside, and I feel guilty for admitting that” (p.175). This statement reflects an internal conflict, however, an examination may reveal that her sentiment is more that of classism rather than racism. Johnson (2006) goes on to explain how parents with higher degrees of wealth are able to “buy-in” to good public schools in good neighborhoods, or opt out of the public school system all together and send their child to a private school. She explores the paradox that exist between the belief in the American Dream where people will be rewarded for their efforts and that anything is possible no matter your background, and the understanding that wealth provides advantages especially in education. Further, those families without wealth had severely limited options (Johnson, 2006). Families she interviewed reported openly that were aware of the advantages they had by having wealth. Her description of the power of the American Dream Ideology may help to further explain the eagerness of many low-income families and parents of traditionally underserved students clinging to the hope that sending their children to the perceived good schools will overcome the disparities created by wealth or the lack thereof. Johnson and Shapiro’s interviews also revealed that many of the white families in the St. Louis suburb they studied feared that as more Blacks moved in there would be 28 increased violence, drug use, lower property values, and an overall deterioration of the neighborhood. Relating this sentiment back to the question of why Black families believe it is worth it to bus their children to other communities leads to another question to consider and the challenging of widely held beliefs. If the sentiment reported by Johnson & Shapiro is pervasive and well known in Black communities, it is then conceivable that financial means is not the only impediment for Black families to move into predominately white neighborhoods. In other words, do Black families know that they are not welcome, and therefore remain in the communities that offer social cohesiveness? This question may never be answered or reconciled. What exists is a paradox. If we accept that all parents, regardless of race and class, want the best for their children, then why would certain families be unwelcome? Further study on race and cultural acceptance may help to answer this. This again highlights the conflict between the social goal of integration and the private goals of acceptance. More specifically, it appears that in the hopes of breaking enduring patterns of racism and social isolation, parents of disadvantaged students are willing to sacrifice and send their children to neighboring schools in the hopes that they realize the educational benefits promised to them (Johnson & Shapiro, 2003; Morris, 1997; Smrekar & Goldring, 1999). Mreover, because Blacks are overrepresented in poor communities and are often characterized as not having the same aspirations as Whites, it is clear to some Blacks that they must work harder and achieve greater in order to overcome these enduring stereotypes. This recognition supports data presented in the later discussion of charter schools showing that charters attract a higher percentage of Black students than traditional public schools, in 29 part because they tend to be located in urban areas. This fact may lead to the following conclusion: When opting for school of choice, in contrast to schools in predominately white suburbs, Black parents will prefer to send their children to schools that are perceived higher quality if those schools are located in urban areas as opposed to schools in the suburbs in order to remain in areas that are culturally familiar. As a result, many of the efforts at trying to improve the K-12 educational outcomes of the underclass in the United States have focused on providing increased school options to Black, Latino, and historically underserved students. Some see school choice as one such mechanism for change. The intent of such choice is to empower parents to gather information and shop around for schools that best satisfy his or her given preferences. Exercising this choice is expected to benefit those that leave an underperforming school and even possibly those that stay behind in that administrators will be forced to reform their education delivery practices or even lose more students and experience drastic declines in enrollment. In contrast to the theory of how school choice will improve all schools, superintendents reported making significant programmatic cuts due to the strain of decreased enrollment and the loss of revenue rather than making dramatic improvements to instruction or the delivery model. Those cuts included cutting art, music, and physical education, eliminating foreign language and science in elementary school, increasing class size, eliminating field trips, a reduction in counseling, and eliminating programs in high school. This is exacerbated when families choose to go outside of the their home district because the money follows the child. Therefore, it stands to reason that urban 30 high-poverty districts will be affected most by declining enrollment due to choice. Those districts, however, that experienced gains in enrollment had the opposite result (Jimerson, 2002). The pattern that emerges is central to those advocating for a market-driven education system. The programs and activities that parents find most appealing and attractive are the ones that are cut when districts face declining enrollment (Fowler, 2002). This exacerbates and speeds up the school’s decline. Any hopes that the school will improve are significantly impaired and those that suffer are the students who are unable to access any of the school choice options presented to them. The time it takes for reform efforts to take hold and to see a corresponding increase in school and faculty efficacy is not always congruent with the high-anxiety and pressure schools face to improve now (Muhammad, 2009). Further longitudinal studies may explore student achievement in both the schools that receive and lose their disproportionate numbers of students through choice programs. Many see school choice as a means for White families to bypass integration efforts and remain suspicious that choice will improve the quality of schools attended by mostly Black and Latino students. Critics point to post Brown v. Board of Education policies of the South. These “freedom of choice” plans were a way for white families to avoid forced integration and continue to deny Blacks, Hispanics, and Native Americans equality of educational opportunity (Farrell, Jr. & Mathews, 1990; George & Farrell, Jr., 1990; Jones-Wilson, Arnez, & Asbury, 1992). It follows then that if the poor cannot move out of their neighborhoods to escape their ineffective schools, as the more affluent have already done (Johnson, 2006), then the school or school district essentially has a 31 monopoly over them. In theory, providing choice can break this monopoly. However, lack of transportation and cultural barriers may result in low income, historically disadvantages, and students of color remaining in their failing schools. Exposure alone to high-performing students, however, may be sufficient to increase the outcomes for below-average students. Bennett and Sani (2008) go on to explore the link between group identification and the depersonalization that takes place when members of the same group begin to use terms like “we” and “us” instead of “I” and “me.” Further, this depersonalization, they argue, causes group members to begin possessing the stereotypical characteristics that are normally assigned to that group. In the case of academics, traits like hard working, involved, motivated, opportunistic, and problem solver may be gradually picked up as new students are assimilated into the new group. Although this study does not attempt to explore the psychological, social, and emotional conditions experienced by students who transfer schools, the belief that students’ achievement will increase through the exposure to high-achieving students is best supported by the theory of cultural ecology. Broadly, it refers to how humans adapt to their physical and social environments in order to survive. For several decades John Ogbu studied the performance of minority children in school and has identified the two sets of factors influencing minority schooling; 1) the educational system, and 2) community forces. In his work with Herbert Simons (Ogbu & Simons, 1998), they use the term minority schools to describe the characteristics under which minorities have been historically educated in the United States. Not only the urban locations of many 32 schools, but the perceptions and cultural characteristics of those doing the educating. Broadly, the educational system is comprised of policies, practices, and reward systems in place that reinforce and define academic achievement. The system is also characterized by how minorities or any subordinate group are treated in schools. Ogbu and Simons found that the denial of opportunity and rewards to minorities characterized school systems (1998). Community forces are the perceptions of minorities and how they respond to the mistreatment they have experienced by whites. Discrimination, denigration of language and culture, policing practices, and unwarranted barriers all influence how minorities respond to education (Ogbu & Simons, 1998). They concluded that even if some African American students possessed the academic ability to succeed, many would not because high achievement was considered “acting white.” As his theory relates to school choice, it is important to consider whether moving minority to students to perceived “good schools” in the suburbs will have a positive effect on their academic achievement. The relationship between school choice and race has yet to be disentangled to a degree that will allow policy makers to implement effective strategies to eliminate the achievement gap between Black and white students. In 2003, Ogbu & Davis studied an affluent suburb in Ohio where white and Black social classes were similar. In their results, they find that although Blacks express high goals, they do not always exhibit or engage in the behaviors that promote academic success. Their findings contradict, to some degree, the self-categorization theory espoused by Bennett and Sani, (2008). In my 33 earlier discussion of enrollment patterns, we found that when exercising choice, Black families are more likely to enroll in schools located in urban areas. Since many school and district leaders base school choice efforts on desegregation and improved outcomes for minority students and the poor, the discussion would not be complete with consideration of how minorities respond to schooling. Changes to the school structure alone do not address the notion that a students’ academic success is impacted by community forces, and how those community forces can contribute to the failure of minority students (Foster, 2004). The discussion of school choice policies related solely on the “system” would be inadequate. If the ultimate goal is the academic success of those groups that have been historically marginalized by society, then the school reform efforts need to take on a broader scope and address the community forces in tandem with “the system.” Ogbu implies that academic achievement of minority students cannot be achieved without both the systemic changes combined with strategies to address the poorly adapted cultural norms that exist in many communities. closure. Further, proponents of open enrollment have only attempted to address the policy and practical factors within the “system” without addressing “community forces” which will have a greater or equal impact on the overall educational outcomes of minority students (Ogbu & Simons, 1998). While race, poverty, and mobility remain powerful contextual explanations in school choice patterns, additional push and pull factors are relevant in school choice decision-making. Parent education, experience, and income may also factor in to decision-making. Further research in this area specifically related to school choice and 34 observed patterns may attempt to answer the question of whether low-income or less educated parents are equipped to make good decisions. Lareau (2002) does however provide insight into how observed class differences impact childrearing. As her work relates to school choice, she notes that reliable transportation and flexible work schedules are necessary for parents to cultivate and transmit the advantages of social class. Furthermore, parents’ ability to gather and process information is essential if they are to avoid making poor decisions (Schneider, Marschall, Teske, & Roch, 1998). Ball and Gewirtz (1998) and Ridenour et al. (2001), divide parents into three categories of choosers: the skilled/privileged, the semi-skilled, and the disconnected: 1. The privileged/skilled chooser is more likely to know how to choose more selective schools because of their substantial personal experiences with schools and social capital. 2. To a lesser extent, the semi-skilled chooser, who has some knowledge but who is less able to determine fiction from fact, may experience subtle personal benefits. 3. The disconnected chooser is unable to critically examine and compare a wide range of schools. It is not due to a lack of interest, it is more often a lack of knowledge (Ridenour, Lasley, II, & Bainbridge, 2001, p. 75). Ridenour, et al.’s theory that there exists less able choosers reinforces the position of those opposed to widespread choice policies because it relies on parents being able to make well-informed choices, but not all parents are the highly informed and engaged consumers that are necessary for them to compete in an education marketplace. Schneider, Marschall, Teske, & Roch (1998) go on to describe four stages that parents 35 go through when deciding what and how to choose: 1) parents must decide on a set of preferences that they want to realize, 2) parents engage in information gathering, 3) make trade-offs between neighborhood school and choice school, and 4) monitor the school’s and their child’s performance to make sure that their choice was a good one. They go on to assert the “disparities in different types of parents’ resources, involvement, and cognitive abilities are particular concerns when considering how parents will play an expanded role in choosing schools” (p. 490). There is little disagreement that more advantaged parents generally have better knowledge about relevant factors in choosing a school (Archbald 2000; Teske and Schneider 2001). What also emerges from this debate is the cultural divide that exists between poor and middle-class families. During the 1990s, a time when school-choice as a public policy grew, many researchers explored the cultural conflicts that exists and compared how families from different backgrounds often seek different outcomes for their children (Cookson 1994; Delpit 1995; Elmore 1991; Fuller, Elmore, and Orfield 1996; Henig 1994; and Hirsch 1996). Several themes emerged from their work. Notably, poor people are as committed to the education of the children as the more economically advantaged. Secondly, in many urban centers, school choice movements have the capacity to tear communities apart and breed a culture of racial resentment and bitterness between those who have exercised choice and those who have not. Next, for a complete understanding of school choice, one must accept that conflict will arise from the practical applications of such polices and the moral imperative to provide all students with access to the best educational options. Lastly, policy changes alone are inadequate 36 in any school reform effort and that without changes, those most at risk and the marginalized are the ones most likely to suffer from a system that collapses. “Educational reform requires a more profound, more lasting, and more structural commitment to families, teachers, and children (Cookson, 1994, p. 11). Delpit (1995) argues that this cultural conflict is most noticeable when contrasting the ideas of white liberal reformers with Black reformers. Whites, who may already be fully participating in the culture of power, advocate for a system that does not arbitrarily force standards upon their children. While Blacks, who have experienced marginalization, “want to ensure that the school provides their children with discourse patterns, interactional styles, and spoken and written language codes that will allow them success in the larger society” (p.29). Similarly, Schneider, et al. (1997) points out that several disparities among parents exist which may impede their ability to become actively involved in their newly expanded role in the school marketplace. These disparities are namely parent resources, involvement, cognitive abilities, and educational experiences. This collection of work is best summarized by Schneider, Marschall, Teske, & Roch (1998) in asserting that “whites and racial minority groups may want different things from the schools, and given choice, greater segregation will result as parents of different class and racial groups choose schools structured on different organizing principles” (p.493). In their research they examinied how families of different social classes and racial groups differ in their evaluation of four school characteristics: 1. the academic quality of the school, 2. the racial composition of its student body, 37 3. the values espoused by the school, and 4. the school disciplinary code. Schneider et al., (1998) sought to discover the degree to which each of these dimensions of school was a function of race and class. They hired a research laboratory to interview 400 families representing the four school districts in the New York metropolitan area. Families must have had a child attending school in any grade K-8 from either private or public school. With regards to academic quality, Schneider et al., found that Black parents and parents who graduated high school, but not college, ranked high test scores as important. Next, parents that had reported attending college were eight percentage points less likely to say that high test scores are important. It follows then that Blacks and parents with no more than a high school diploma equate high academic quality with test scores more so than middle and upper class parents with more education. As more minority and historically marginalized groups have increased awareness around the requirements for academic success, they will be demanding more of their schools or seeking other schools through choice programs (Delpit, 1995). The second finding was that white parents and parents with education levels of some college or above were more likely to rank the school’s values as important than families of racial minorities and lower education levels. On the third dimension, discipline was valued much less by white parents and parents with higher education than minority parents and parents with less education (Schneider, Marschall, Teske, & Roch, 1998). Table 2.1 shows how each group ranked the four different school attributes. 38 Table 2.1. School Attribute Rankings of 1,582 Parents in New York City and Suburbs Black Asian Hispanic H.S. Graduate College High Scores 1 2 2 1 4 Values 3 3 4 4 1 Discipline 2 1 1 2 3 Diversity 4 4 3 3 2 (Schneider, Marschall, Teske, & Roch, 1998) Note: See Appendix A for complete table of the change in probability. Charter Schools As previously described, race, culture, and economics are strong factors to consider when examining school choice. With regard to the use of charter schools as a specific form of school choice, the data suggest a different trend. According to the University of California Civil Rights Project, in their report Choice without Equity: Charter School Segregation and the Need for Civil Rights Standards, several important facts regarding race and charter schools are noted: Latinos are disproportionately less likely to enroll in charter schools in five of the six states with the largest shares of Latino students. Nationally, white students are enrolled in charter schools at a rate lower than their representation, but in the U.S. West, where traditional public schools are the most racially diverse whites are over-enrolled in charter schools. Charters attract a higher percentage of Black students than traditional public schools, in part because they tend to be located in urban areas where more Black students may tend to live (See Table 2.2). 39 In the industrial Midwest, more students enroll in charter schools compared to other regions, and Midwestern charter programs display high concentrations of Black students (The Civil Rights Project, 2010). Table 2.2. Black Over-Representation in Charter Schools, top 15 MSAs, 2007-2008 Metropolitan Statistical Area (MSA) Charter School Black % Public School Black % Black Charter OverRepresentation Cincinnati-Middletown, OH-KY-IN 78% 15% 520.0% Kansas City, MO-KS 79% 17% 464.7% Dayton, OH 74% 17% 435.3% Boston-Cambridge, Quincy, MA-NH 32% 8% 400.0% New York-Northern New Jersey-Long Island, NY-NJ-PA 66% 20% 330.0% Indianapolis-Carmel, IN 62% 19% 326.3% Minneapolis-St. Paul-Bloomington, MNWI 37% 12% 308.3% Chicago-Naperville-Joliet, IL-IN-WI 64% 22% 290.90% Washington-Arlington-Alexandria, DCVA-MD-WV 88% 31% 283.9% Detroit-Warren-Livonia, MI 72% 26% 276.9% Cleveland-Elyria-Mentor, OH 70% 27% 259.3% San Antonio, TX 17% 7% 242.9% Los Angeles-Long Beach-Santa Ana, CA 17% 7% 242.9% Milwaukee-Waukesha-West Allis, WI 50% 23% 217.4% Columbus, OH 42% 20% 210.0% Source: Frankenberg, Siegel-Hawley, & Wang, 2010 An examination of charter school applicants to New York City charter schools reveals a similar trend. Table 2.3 shows that charter school applicants were largely Black and Hispanic, 63% and 29% respectively. Only a few percent were white, Asian, or 40 another race. Hoxby (2009) notes that all of the charter schools in New York City are oversubscribed and use a lottery system for admission. There will be further discussion of the results of this study in the next section covering Effects on Student Learning. Table 2.3. The Race, and Gender of New York City Charter School Applicants and Students in Traditional Public Schools from 2000 to 2006 Students' Race / Ethnicity Black non-Hispanic White non-Hispanic Hispanic Asian Other race All Applicants to charter schools Applicants who were lotteried-in Applicants who enrolled in charter schools New York City's traditional public schools 63% 4% 29% 3% <1% 64% 4% 28% 3% <1% 61% 4% 29% 4% <1% 34% 15% 38% 12% <1% (Hoxby, The promise and performance of charter schools: Drivers of educational improvement in the U.S.?, 2009) Table 2.3 also shows that charter school students in New York were disproportionately Black and Hispanic with only a few percent white, Asian, or of other race. The table also shows that compared to all applicants, those students that were lotteried-in are nearly identical to those lotteried-out, indicating that the selection of enrollees was truly random. As a result, New York’s traditional public schools are far less Black, and more white and Asian than charter schools in New York. Mathematically, there are nearly four times more Asian and white students in New York’s traditional public schools than in charter schools. From this data, it appears that Black students and their families use choice options much more than others do. What the data does not answer is why. This study addresses this question. 41 Years of research on social and economic isolation has linked segregation to higher crime rates, unemployment, lower academic achievement, more aggressive policing policies, and teenage childbearing (Bursik 1998; Lauen 2007; and Harding 2003). Additionally, inner-city and geographically isolated residents have decreased access to information regarding jobs, education, and opportunities for mobility. Without adequate social and financial resources, families in disadvantaged neighborhoods are less likely to seek, apply for, and travel to schools of choice. As a result, many of the reform efforts have focused on providing increased school options to Black, Latino, and historically underserved students. School transfers and choice are seen by some as one such mechanism for change. This proposed shift in power is intended to empower parents to gather information and shop around for schools that best satisfy their given preferences. Many, however, see school choice as a means of bypassing integration efforts and are suspicious that choice will in fact improve the quality of schools attended by mostly Black and Latino students. Critics point to postBrown v. Board of Education policies of the South. These “freedom of choice” plans were a way for white families to avoid forced integration and continue to deny Blacks, Hispanics, and Native Americans equality of educational opportunity (Farrell, Jr. & Mathews, 1990; George & Farrell, Jr., 1990; Jones-Wilson, Arnez, & Asbury, 1992). Effects of Choice on Student Learning Research in the area of school choice has sought to examine the effects choice has on the educational outcomes of students. More specifically, the research has sought to 42 answer the question: does school choice increase student achievement at both the receiving schools and the schools that students are leaving? Low-income, minority students in urban areas tend to be the focus of open enrollment and school choice efforts. Advocates claim that these students, and others from racially and economically isolated schools, are the ones that will see the biggest benefit from attending higher performing, more integrated schools (Dillon, 2008). Extending the market effects analogy to K-12 education, Ledwith (2010) points specifically to research conducted by Dee (1998) and Hoxby (2000) where graduation rates and student achievement increased subsequent to enacting open enrollment policies. The following are the main findings of Hoxby’s analysis of New York City charter schools: On average, a student who attended a charter school for all of grades K-8 would close about 86 percent of the achievement gap in math and 66 percent of the achievement gap in English. A student who attended fewer grades would improve by a commensurately smaller amount. On average, his lotteried-out counterpart who stayed in the traditional public schools for all of grades K-8 would stay on grade level but would not close the achievement gap by much. However, the lotteried-out students' performance does improve and is better than the norm in the U.S. where, as a rule, disadvantaged students fall further behind as they age. 43 Compared to his lotteried-out counterparts, a student who attends a charter high school has Regents examination scores that are about three points higher for each year he spends in the charter school before taking the test. A student who attends a charter high school is about seven percent more likely to earn a Regents diploma by age twenty for each year he spends in that school. For instance, a student who spent grades ten through twelve in charter high school would have about a twenty-one percent higher probability of getting a Regents diploma (Hoxby & Murarka, 2009, pp. IV-I). An important fact of this study and similar studies that compared lotteried-in and lotteried-out is that both cohorts of parents participated in the lottery and therefore it can be concluded that they both have similar levels of parent involvement. This directly refutes the claim that uses parent involvement as the deciding factor in student achievement. Effective evaluation of school choice programs suffers from many limitations. Often times controlling for demographic differences and selection biases are hurdles. However, changes in testing formats, other school improvement strategies, residential decisions by parents, application procedures, prior student achievement, and availability of transportation are additional factors that make measuring the effectiveness of particular programs difficult and often inconclusive. The criteria used in the selection of students for school choice programs provide a way of separating the effect of the program from the effect of other characteristics. Furthermore, since most choice programs are 44 oversubscribed, those students that gain admission and those students that do not, make natural treatment and control groups (Ballou, Goldring, & Liu, 2006, ). Miron, Evergreen, & Urschel (2008), in a policy brief to examine and summarize the evidence regarding school choice and its impact on student achievement, examine 87 school choice studies to measure the effect on student achievement. The authors based their selection of these studies on the following seven criteria. 1. Presence of the technical report with a clear explanation of the procedures used 2. Presence of analysis and conclusions 3. Use of standardized test scores to measure student achievement 4. Use of comparison groups 5. Exclusion of duplicated studies 6. Exclusion of case studies or single school studies 7. Exclusion of studies on school choice outside the United States Of the 87 studies examined, nine were related to inter-, intra-district, or magnet school programs. For the purposes of this dissertation, I will only look at these nine studies. The authors analyzed each study and gave ratings on the quality of the study, and impact on student achievement. Positive values for impact indicate that the choice program showed an increase in student achievement. A scale value of 2 indicated a positive overall effect, 1, a slightly positive effect, 0 mixed impact, -1 slightly negative overall impact, and -2 negative overall impact. Secondly, the studies were given a quality of study score from 0 to 32 based on six different dimensions including research design, duration of study, controls, measures of student performance, scope of study, and completeness of technical report. 45 The results illustrated in Figure 2.1 show that only four had a “slightly positive” effect on student achievement with none demonstrating a “very positive” effect. Moreover, two of the nine programs showed a “slightly negative” or “very negative” effect on student achievement (Miron, Evergreen, & Urschel, 2008). Figure 2.1. Quality and Impact Ratings for Studies of Student Achievement in Magnet or Interdistrict Choice Schools (Miron, Evergreen, & Urschel, 2008) Leading with the highest quality studies first, Study H, Does School Choice Work? Effects on Student Integration and Achievement, (Betts, Rice, Zau, Tang, & Koedel, 2006) compares lottery winners and losers in an urban intra-district choice program. The San Diego Voluntary Ethnic Enrollment Program (VEEP), which started after the court-order desegregation of schools in the 1970s, is another example of a program that sought to increase the racial diversity at San Diego schools and increase student achievement. Despite achieving the type of integration that open-enrollment advocates argue for, there was little evidence to indicate students where achieving higher degrees of academic success. Betts et al., examine the achievement of students after San 46 Diego implemented their VEEP program. The two groups were labeled lottery-winners and lottery-losers. This is similar to what Hoxby describes as lotteried-in and lotteriedout. Betts and colleagues’ findings indicate that there was no significant difference in math and reading achievement one, two, or three years after the lottery took place. Surprisingly, they found that after the first year of transfer, there were temporary loses among the lottery winners. (Betts, Rice, Zau, Tang, & Koedel, 2006). This last discovery opens the door to exploring the disadvantages of non-routine change of schools. When students transfer schools at times other than from elementary to middle school, and middle school to high school, they are at risk of mitigating any advantages they would otherwise enjoy by attending schools that follow a sequenced course-taking pattern (Schneider, Swanson, & Riegle-Crumb, 1998). Increased residential mobility and frequent school transfers is often associated with negative social outcomes, higher drop-out rates, more serious discipline problems, and lower level of educational attainment. Moreover, students who make non-routine school changes are often misplaced in their new school’s curricular structure. Schneider, et al. describe this as curriculum dislocation, which places them at a disadvantage to those students who have not had non-routine school changes (Schneider, Swanson, & Riegle-Crumb, 1998). Further research on this topic may explore how other factors of adolescence, family influences, and residential moves affect student achievement during non-routine changes of schools. Study B, Magnet Schools and Student Achievement, (Ballou, Goldring, & Liu, 2006) compared lottery winners and losers and makes up for deficiencies in other studies 47 that do not control for individual student differences and changes in the demographic composition of the schools. The seven variables controlled for were race, gender, English proficiency, disability, poverty, and previous grade’s reading and math scores. This study examined the effect of magnet schools in a mid-sized Southern city. The district serves students in grades K-12 and is racially mixed with 40% White students and 48% Black. The focus of this study is on middle school students who had applied to one or more of the district’s magnet programs through lottery. The districts’ magnet program showed mixed results. There appeared to be a positive impact on magnet students’ math scores, however, when prior levels of achievement and demographics were controlled for, the results in math mirrored the district as a whole. Additionally, Ballou et al., (2006) found that although there appeared to be a positive magnet effect on math scores, the growth was not sustained over time as would be expected with the additional benefit of attending a magnet school. In contrast, Ballou et al., (2006) report that there is less evidence that the magnet school is having a positive effect on reading scores. Similar to math, there was no evidence that students would experience a cumulative academic gain by attending a magnet school for consecutive years. Another important factor uncovered by Ballou et al., (2006) is the role of attrition when trying to compare lottery winners and losers. Their analysis of attrition showed an interesting pattern. Lottery losers with lower test scores relative the lottery winners left the district with greater frequency than lottery winners with similar levels of achievement. It appeared that lottery winners, regardless of level of achievement were 48 confident that they found a satisfactory schooling option and were likely to stay in the district. Following, if low test scores were the motivation for families to consider different schooling options, they will continue looking if they fail to win a lottery. This means that the control group of lottery losers will lose more of its weaker students that the group of lottery winners, the treatment group. The implication is that measuring the effectiveness of magnet schools becomes more difficult without controlling for attrition and previous levels of achievement. Study D, Student Achievement in Public Magnet, Public Comprehensive, and Private City High Schools (Gamoran, 1996), and Study E, The Impact of Career Magnet High Schools: Experimental and Qualitative (Heebner, 1995), both study magnet high schools that have a career focus. Gamoran (1996) uses National Educational Longitudinal Survey (NELS) test data from 1988 – 1990 to compare the gains in student achievement from 8th to 10th grade for magnet schools, public comprehensive schools, and private Catholic schools. In general, he found that magnet schools are more effective than regular public schools at raising student achievement in science, reading, and social studies, and when controlling for preexisting levels of student achievement non-secular private schools did not offer any advantage. Consistent with social and economic advantages, students at private Catholic schools scored higher in math, science, reading, and social science. The results of his study will favor more and support those who advocate for more specialized programs in public schools. It also appeared that the magnet schools studied served more disadvantaged students than did the regular comprehensive schools. This 49 trend may reflect the growing outcry for reform of inner-city schools and the resulting efforts for school transformation. Additionally, as more districts expand their school choice options, the growth in urban magnet schools may be an effort to attract students from outside the boundaries, and retain those from within who may otherwise leave. Heebner (1995) uses data from five career magnet schools in New York to examine the outcomes for students who gained admission through lottery and those that gained admission through a selection process. Because of the uniqueness of the program and the means for selection of the participants, the results may not be generalizable to other groups. However, lottery winners did experience higher math scores than those attending regular public schools. In reading, students with average reading scores demonstrated more growth than their counterparts in regular public schools (Heebner, 1995). Study G, Evaluation of the Magnet Schools Assistance Program (Christensen, Eaton, Garet, Miller, Hikawa, & DuBois, 2003), is a report that examines the progress MSAP projects have made in meeting the goal of improving educational outcomes for minority students and reducing minority student isolation. The report pointed out that each school was required to establish goals in various domains that did not always include language arts and math. Final goals were a set target level of proficiency at the end of the grant period. While all of the schools had goals, only about half of the MSAP schools had data that was analyzable. For example, schools had goals for educational attainment, preparation for college, or alternative assessments. Nonetheless, Figure 2.2 shows the percentages of schools that met at least half of their final goals. 50 Figure 2.2. Percentage of Schools that Meet Half or More of Their Final Goals (Christensen, Eaton, Garet, Miller, Hikawa, & DuBois, 2003) Concerning the stated purposes of MSAP, 57% of the grant funded schools succeeded in desegregation efforts when adjusting demographics to the current district trends. Conversely, 43% of the schools were not successful in eliminating or reducing minority group isolation (Christensen, Eaton, Garet, Miller, Hikawa, & DuBois, 2003). Next, MSAP supported schools were most successful during their first year of implementation in meeting their goals, but over longer periods of time the gains were not fully sustained. When addressing the third and fourth purposes of innovation, improved methods, and systemic reform, MSAP supported schools had adopted reform models at greater frequencies than non-magnet schools. Additionally, survey data indicated tat MSAP schools had a more positive school climate and teachers reported a greater emphasis on higher level thinking skills. However, the case studies highlighted tensions that sometimes existed when trying to implement systemic reforms and innovative instructional practices (Christensen, Eaton, Garet, Miller, Hikawa, & DuBois, 2003). 51 Overall, the results showed that the most relevant school factor improving student achievement was school climate. Christensen et al., report that schools with more positive program features outperformed comparison schools. However, most MSAP schools had adopted innovative themes and had somewhat more positive climates and professional learning communities than did comparable schools. In addition to a comprehensive examination of varied school choice programs, the work of Miron et al., demonstrates that well conceived school choice initiatives that are effectively evaluated and studies have a more positive impact on student achievement. Four of the five studies that received a quality of study score of 19 or higher also showed at least a slightly positive effect on student achievement. Conversely, those studies that received a quality of study score of 17 or lower either had no effect or a negative effect on student achievement. Study C, The Effects of Academic Career magnet Education on High Schools and Their Graduates, (Crain, et al., 1999), studied career magnet programs in large lowincome cities with high populations of African American or Hispanic students. Their study consisted of two parts. First, a quantitative analysis of 9,176 student data files compared test performance, attendance, graduation rates, and dropout rates of lottery winners and losers to career magnet programs. What their analysis uncovered was that many career magnet programs have lower graduation rates and higher dropout rates than comprehensive high schools. Higher academic standards and a tendency to push weaker students out seemed to be a cause. In addition, admission to these programs is not solely by lottery; up to half of the students were handpicked. Subsequently, students would 52 have a greater chance of graduating had they lost the lottery. The analysis of student data files revealed that students in career magnet programs did not have significantly different math and reading scores, nor do they take advanced graduation tests at rates different from their counterparts in comprehensive high schools. Second, 110 high school graduate lottery winners and losers were surveyed and interviewed. In addition, the researchers selected 30 participants from the survey respondents for four-hour interviews. Their analysis showed that graduates from career magnets were more likely to have chosen a major within the first two years of college attendance and have earned more college credits. The additional positive effects related more to the psychological and social benefits small learning communities. Career magnet graduates reported they smoked less, had fewer fights, drank alcohol much less often, and became pregnant or caused a pregnancy less often (Crain, et al., 1999). This last observation supports the position of the open-enrollment proponents who subscribe to Ogbu’s theory of cultural ecology. The magnet schools appear to better help students through their identity development process and adolescence. School attachment, perceived teacher support, parental control, and minimal exposure to community violence are linked to greater student outcomes and lower levels of delinquency (Frey, Ruchkin, Martin, & Schwab-Stone, 2009). Study A, (Beaudin, 2003) was a report to the Connecticut Department of Education on interdistrict magnet and magnet schools in the state. It compared the scores of students in magnet schools with the statewide averages over a two-year period. The results were mixed and showed both positive and negative effects on the two different 53 standardized tests. Additionally, the study’s lack of controls for demographics or other reform efforts that individual schools were undertaking accounted for the low rating in quality of study (Beaudin, 2003). During the two years studied, 2000 and 2001, students enrolled in magnet schools in grades 4, 6, and 8 in Connecticut scored below the state average and below students from high socioeconomic backgrounds. The only group that magnet school students performed better than are students from high-poverty schools. In reading the results are similar. Magnet school students in grades 4, 6, and 8 scored below the state average and were outperformed by students from schools that were predominately white and had low numbers of students in poverty. On the CMT Writing Test, it appears that 8th grade magnet school students are better prepared in writing and outperformed the state average by 4 percentage points in 2001. However, those same students performed 20 percentage points below students in low-poverty schools. Grade configurations for Connecticut schools include K-8th grade and K-6th grade. In examining the data, 6th and 8th grade students are included in both sets, but are not disaggregated any further. Figures 2.3, 2.4, and 2.5 show the comparisons of each group on the CMT Mathematics, Reading, and Writing Tests administered in 2000 and 2001. 54 Figure 2.3. Comparisons of the 4th Grade Students Performing at or Above Connecticut Goal on CMT Writing Test (Beaudin, 2003) Figure 2.4. Comparisons of the 6th Grade Students Performing at or Above Connecticut Goal on CMT Writing Test (Beaudin, 2003) Figure 2.5. Comparisons of the 8th Grade Students Performing at or Above Connecticut Goal on CMT Writing Test (Beaudin, 2003) The overall conclusions from the Connecticut study of the impact of elementary and middle magnet schools on academic performance is mixed. Without controlling for 55 demographics it is unclear if high poverty and students of color are overly enrolled in Connecticut’s magnet schools. If they are in fact attending magnet schools at high rates, then it can be concluded that magnet schools are increasing student achievement and reducing the gap in performance for students who attend them. To date, studies on school choice have primarily focused on the family as the agent empowered with choice. A growing area of research is focusing on the school as the choice agents. When certain market conditions exist, school leaders use various tools to seek and retain students that will enhance their test scores and reputation (Jennings 2010; Lauen 2007). Further research may examine how school leaders use overt and covert means to influence enrollment patterns. Marketing, interviews, test scores, family and child demographics, and the social network that exists among principals are among some of the ways schools may weed out those students that require higher levels of service including those that are English learners, have special needs, or have behavior problems (Jennings 2010; Lauen 2007). Skimming refers to the practice of enrolling high ability students and leaving out lower achievers and students with disabilities. Although lottery-based studies do not support that charter schools in fact do intentionally “skim” the top students from regular public schools, with regards to race and economic status, there is substantial evidence to suggest that white parents and parents with higher socioeconomic status exercise other school choice options at higher rates than do Black, Latino, and economically disadvantage families (Fisk and Ladd 2000; Hsieh & Urquiola, 2006; Teske and Schneider 2001). Zimmer et al., (2009b) studied locations Chicago, Denver, Milwaukee, 56 Ohio, Philadelphia San Diego, and Texas examining student level data for stratification. Their findings did not show that charter schools systematically skim high achieving students from regular public schools. This may lead to the conclusion that open enrollment polices do nothing to close the achievement gap among white students and students of color. Impact of Choice on Non-Choice Schools Another undeniable aspect of the effects of school choice is the impact it has on the schools that are the net-losers of such policies. What is more compelling is the influence of culture on school choice decisions. While some will overemphasize the extent to which low-income families will disengage from the process, “social norms and ethnic identity constrain[s] the range of school choices that are credible in their eyes” (Wells, 1996, p. 25). In addition, the lack of adults willing or capable of advocating for those students will ultimately relegate them to the inferior school even when choice exists (Ridenour, Lasley, II, & Bainbridge, 2001). Ledwith (2010) goes on to acknowledge the points made by opponents of school choice. She points to the work of Fiske and Ladd (2000), who argue that families without adequate financial means are not able to take advantage of choice options. Further, she states that students and families “will not be in a position to choose good schools and will become trapped in ‘hollowed-out’ public schools” (p. 245). Moreover, she points to the “skimming” that can occur of white and non-Hispanic students from traditional public schools. These “skimming” practices are usually attributed to charter schools in which selection, enrollment, and retention 57 practices attempt to include only students with greater potential for academic achievement. Continued Debate on School Choice Does school choice matter in the achievement of students and whose parents chose to exercise it? This is the very important question that policy makers wish to answer. An argument used in support of school choice is that the competition it generates will significantly raise the quality of public schools. The market analogy to consider is if you could only get your auto serviced by one mechanic (your child can only go to one school). Granting such a monopoly provision would cause most to expect that the quality of service provided by this one mechanic at a given price would be far less than if the consumer had many auto mechanics to chose from to service their auto. Supporters go on to argue that school choice will motivate public schools to be more responsive to parent and student demand and promote the pedagogical changes necessary to improve student outcomes (Tooley 1993; Whitty & Edwards 1998). Additionally, competition through open enrollment will create incentives for traditional public schools to embrace reform. An Indiana study, Study of the Effectiveness and Efficiency of Charter Schools in Indiana, conducted by the Center for Evaluation and Education Policy, states, “it appears that charter schools have played some role, through market competition, in motivating school[s] to make positive structural and programmatic changes” (Akey, et al., 2008, p. 105). The study, however, did not note 58 any significant improvements in student achievement, but it illustrates that when schools and districts are faced with potential declining enrollment because of choice options, choice can be a catalyst for organizational change. The report quotes one principal as saying: “Some really great ideas are emerging. Ultimately, you have this innovative culture. People are always thinking about how to differentiate ourselves and meet the needs of our children. We’re seeing significant changes from a climate where everything has to be the same. The others run with it and say ‘we can do it too.’ There are different treatments for different needs. That’s where I see different ways to approach this work” (Akey, et al., 2008, pp. 105-106). The data for the Indiana study came from three major sources. First, CEEP obtained the qualitative data related to student achievement, attendance, and graduation rates from the Indiana Department of Education and individual school accountability reports. Qualitative data was obtained through interviews with 30 community stakeholders. They included superintendents, IDOE staff, leaders from professional organizations, local university staff, and staff from the city mayor’s office. The CEEP collected additional data from documents related to charter school contracts and procedural policies. The transformation of local public schools to become market driven institutions requires the acceptance of certain assumptions related to both schools and the teachers 59 within them. First, these schools must not already be market-driven. Second, school systems would now offer educational options that matched the demands of parents and students. Third, limited mobility for families in poverty and private schools being cost prohibitive strengthens the monopoly of the local public school. Lastly teachers would embrace the accompanying changes and make corresponding changes to their instructional practices. This last point assumes that the teachers have not already adopted the changes in instructional practices. In a system that has operated as a monopoly for decades, the challenge becomes not only changing the system, but also changing the people within it. Gaining market share in schooling equates to increased enrollment, it offers prestige to those teaching there, and it provides power and influence in the system (Smrekar & Goldring, 1999). The prestige that comes with the knowledge that your school is preferred to others, offers greater empowerment to teachers and may further increase teachers perceived efficacy. Furthermore, this empowerment has a tremendous impact on satisfaction when it is within an organization that supports the individual by providing the vision, training, and resources necessary (Bogler & Nir, 2012; Fullan 2008). Bogler & Nir (2012) go on to report that the intrinsic satisfaction that comes from empowerment is the self-efficacy one experiences when their psychological and social needs are met. Steffens and Cookson, 2002, however, make the claim that the market metaphor does not apply to public education. “Public education is a social commitment that transcends individual interest and corporate gain…[A]s a human service, education is grounded in a belief in human dignity that transcends the values and behaviors 60 associated with markets. It means public education cannot be squeezed to fit the market model and still meet the needs of a just society” (p.5). Similar to public education, health care is another commodity that is provided in market systems. Beyond improving academic achievement and the other previously described outcomes, proponents of school choice contend that open enrollment will have a positive effect on reducing racial and economic segregation in public schools (Dillon, 2008). The UCLA Civil Rights Project reports that 40% of Black students in America attend a school that is at least 90% Black. That number is up from 1988 figures that showed that onethird of Blacks attended similarly comprised schools. The report goes on to confirm that Black and Latino children were more segregated in 2009 than at the time of Martin Luther King Jr.’s death. This new trend of re-segregation was brought on by a disturbing housing trend that drastically reduced the amount of affordable housing combined with stagnation and slight decline in average wages for American workers has created new concentrations of poverty in America’s cities and schools. With greater attention given to equity issues related to choice, the need to study school enrollment practices has caused The University of California, Civil Rights Project, The Journal of Negro Education, The Journal of School Choice, The Brookings Institute and others to examine whether or not charter schools provide choice without equity. Nearly two decades of study around charter schools led to the following three outcomes: 1. Most states provide incentives for charter schools to serve low-income and minority students, 61 2. The people seeking to run charter schools were predominately dedicated to improving school options for the most disadvantaged students, 3. An increasing number of foundations provide targeted financial assistance to charter schools serving poor minority students in urban districts that have been slow to adopt reforms (Hill & Lake 2010, Nathan 1996, Hoxby, 2009). As previously described, the California Open Enrollment Act was intended to improve student achievement and enhance parental choice in education by providing additional options to pupils to enroll in public schools throughout the state without regard to the residence of their parents. Dillon (2008) states that due to lack of resources, political barriers, and underfunded policies, achievement gains have been inconsistent and can actually exacerbate segregation and increase the achievement gap among Black and Latino students, and white and Asian students. This happens when school choice plans are poorly designed and implemented, and do not target those students that could benefit most from increasing their access to higher performing schools (2008, p.3). Conclusion In attempting to understand the impact of school choice on student achievement, we have yet to collectively define what student achievement means, what factors contribute to it, and how it can be measured. The measures currently in place, IQ tests, standardized test scores, graduation rates, retention and drop-out rates, and college acceptance may be too simplistic to measure complex human behaviors (Cookson, 1994). Additionally, some consider job skill preparation, civic engagement, an appreciation of 62 the arts, character development, physical development, and the appreciation of lifelong learning as fundamental measures of student achievement (Goodman & Zimerman, 2000). Cookson (1994) describes how statistical control remains a fundamental issue in measuring and interpreting data. For example, although school governance has shown to have positive effects on student achievement (Goodman & Zimerman, 2000), so has student family backgrounds, socioeconomics, community forces, and the social, psychological, or religious forces that influence parents to sacrifice for the sake of the child’s education (Cookson, 1994; Delpit, 1995; Lauen, 2007). As described earlier, among those studies that did have suitable controls, some found positive effects of school choice, some found negative effects, and others found no significant effect at all. The researched survey here provides no conclusive evidence that competition through school choice provided any substantial benefit to those students choosing choice schools or those that remained. It highlights the clear need for a deeper analysis and possibly case study research within states or even school districts specifically following cohorts of students that remain within their same school district. The absence of any definitive conclusion on overall student achievement does not rule out the other possible benefits such as other improved outcomes for active choosers, a better match between families and schools, opportunities to participate in unique programs, and the benefits provided by social cohesion (Levin, 1998; Levin, 2002). The mixed results of choice on academic achievement indicate that the issue of school choice is complex. An explanation may be found by further examining more 63 demographic information including suburban vs. urban communities, the number of schools families have to choose from, geographic features, availability of public transit, and race and class tensions. Only recently have school choice policies been implemented on a scale large enough to allow for meaningful study. Although the base of knowledge is growing, the results are inconclusive and can neither support nor refute claims that school choice plans improve the educational outcomes of those students that leave or remain at their neighborhood schools. The absence of any strong evidence confirming that choice polices do improve schools, does not rule out any possible gains in equity, social cohesion, or improved outcomes for those who actively choose (Arsen & Ni, 2008). Researchers have underscored the need for policy makers and school officials to thoroughly understand the dynamics of school choice by making the argument that successful policies cannot be implemented without a thorough theoretical understanding of the traditional school system and the implications to families that choose to go outside that system (Koedel, Betts, Rice, & Zau, 2009). 64 Chapter 3 Methodology Introduction The following chapter reviews the research design for this study. It presents a rationale for the use of a regression analysis to analyze the demographic data of both participants and non-participants in open enrollment, and the use of Pearson correlation coefficients to analyze the results of parent surveys. Both are integral components in understanding the phenomenon of parents choosing to enroll their child in a school other than their home school during open enrollment. This chapter reviews the data collection procedures, describes the study setting, population, and research sample. Chapter 3 also reexamines the research questions and explains how the quantitative data was collected and analyzed. Finally, I discuss the roll of the researcher and the measures used to protect the study participants. Research Questions 1. What variables (available in standard data files) exert a significant impact on the choice to participate in open enrollment, both in terms of them having a non-zero influence and the magnitude of this influence? 2. Through correlation and mean comparison, what factors are significant in relation to why families choose to participate in open enrollment and leave their home/neighborhood school? 65 3. To what degree do current school policies influence open enrollment and school choice decisions? Population and Setting The setting for this study is a large urban and suburban comprehensive pre-K-12 school district in northern California. For the purposes of this study, the district will be known as Golden Unified School District, GUSD, to conceal and protect the identities of the study participants. GUSD serves approximately 47,000 students in grades pre-K through 12. Figures 3.1, 3.2, 3.3, and 3.4 show the demographic breakdown of the student body and academic progress of GUSD over time. Figure 3.1. District Racial and Ethnic Distribution 2010 -2011 School Year Source: Ed-Data, 2012 66 Figure 3.2. District Students Percent at or Above Proficient – Math 2008 - 2012 Source: California Department of Education, 2012 Figure 3.3. District Students Percent at or Above Proficient - English-Language Arts 2008 - 2012 Source: California Department of Education, 2012 67 GUSD is presently in program improvement status. School districts identified as in program improvement status have not met adequate yearly progress in ELA or Math for all numerically significant subgroups for two consecutive years. Not until 2010 did GUSD’s performance fall below the state’s targets. The low initial targets were established with the intent that all students will be proficient by the year 2014. The California Department of Education reports that there are 486 school districts or LEAs Local Education agencies, in program improvement status. That number represents 49% of the 994 districts in the state of California. Schools or school districts can exit PI status when they make adequate yearly progress for two consecutive years. Data I collected quantitative descriptive data on 42,228 students from GUSD for the 2012 – 2013 academic school year. After receiving university and district approval to conduct this research, I met with the GUSD’s research and evaluation department and was provided with an electronic database containing descriptive data described below on every student minus names. Student ID numbers remained in the file so that I could perform a random sampling for the survey data analysis. In addition to indicating which students and are currently participating in open enrollment, the data also included the following: a. Neighborhood school – the school the student is assigned to based on home address; b. Chosen school – the school the student is assigned to based on request; 68 c. Age of child – current age at time of study; d. Grade of child – grade during 2012 – 2013 school year; e. Home language – language other than English spoken at home; f. Primary language – primary language (Table 3.1) the student communicates in; g. Ethnicity (Table 3.2); h. Gender; i. English proficiency (Table 3.3); j. Socio-economic status – low or not low, families that qualify for free or reduced lunch or have a parent that did not graduate high school meet the district’s criteria for Low SES. The following Tables referenced above further illustrate the ethnic and linguistic diversity of GUSD. Among all students in grades 9 -12, thirty-four languages are spoken with a degree of fluency ranging from limited English proficiency to fluent. Table 3.4 shows the total number and percent of students at each overall performance level. Table 3.1. Primary Languages of Students Within GUSD Arabic Farsi Ilocano Polish Armenian Filipino Indonesian Punjabi Assyrian French Italian Romanian Burmese Greek Japanese Russian Cantonese Hindi Korean Samoan Dutch Hmong Mandarin Serbo-Croatian English Hungarian Marathi Spanish Source: GUSD data files Thai Tigrinya Turkish Ukrainian Urdu Vietnamese 69 Table 3.2. Ethnic and Racial Categories of Students Within GUSD American Indian / Alaskan Native Filipino Asian-Chinese Hispanic or Latino Asian-Korean Pacific Islander-Other Asian-other Asian Pacific Islander-Samoan Asian-Vietnamese White Black / African American Source: GUSD data files Table 3.3. Student English Proficiency Designations of Students Within GUSD English Only Initially fluent English proficiency (at time of first testing) Reclassified fluent English proficiency Beginning Early Intermediate Intermediate Early Advanced Advanced Source: GUSD data files The study population included students and families living within the GUSD’s boundaries, or intra-district transfers. This study does not examine inter-district transfers, or transfers of students coming from outside of the district’s boundaries. Additionally, students enrolled in any of the district’s special education centers, alternative schools, charter schools, or home school were omitted from the study. In my analysis, the excluded category of race/ethnicity is White, with the exception of students that identify as White and speak a language other than English at home. Note that the race/ethnicity category Other Asian is included in this school district’s standard student data to account for any other Asian ethnicity not specifically listed on the questionnaire. The home language is other than English category applies to all student regardless of their 70 race/ethnicity and includes students that identify as White, and speak a language other than English at home, including Albanian, Arabic, Armenia, Assyrian, Dutch, Farsi, French, German, Greek, Hebrew, Hindi, Hungarian, Italian, Polish, Portuguese, Romania, Russian, Serbo-Croatian, Turkish, and Ukrainian. The district’s open enrollment policy states that parents of students may apply to enroll their child in any grade appropriate school and are willing to provide their own transportation. A computerized random selection determined which students are accepted for their choice school. Further, the district stipulates that you must participate in open enrollment if: You would like your student to attend a school other than your current school of residence. You would like your student who currently attends a K-6 or K-8 school to attend 6th grade at a middle school. (all district middle schools are grades 6-8) You would like your child to attend a middle school other than their resident K-8 school for 6th, 7th, or 8th grade. Your student qualifies for the High Achiever Program and you want him/her to attend a middle school other their resident school. You would like your child to attend any of the districts six elementary schools, one K-8 school, one 6-8 middle school, or one 9-12 high school that enroll exclusively through open enrollment. 71 Your child was overloaded to the school they currently attend, you have not filled out an intradistrict transfer, and you would like for him/her to remain at the overloaded site. Quantitative Methods Used in this Study I used a quantitative approach in analyzing two sets of data. Within the frameworks of rational choice theory and social cognitive theory, I examined the actual behavior and perceptions of parents regarding open enrollment and school choice decisions. This quantitative approach purposefully used two forms of analyses. First, I used logistic regression analyses of socio-economic characteristics available in commonly collected data by the school district to determine which characteristics correlate with a parent’s likelihood of leaving their home school. This analysis will show the likelihood increase of leaving a school if an included explanatory variable changes by one unit while holding all other explanatory variables constant. For dummy explanatory variables this is a change from not having the measured characteristic to having it. The second form of quantitative analyses used in this dissertation study is the generation of Pearson correlation coefficients and a measure of means to measure how well the variables are related, and to uncover why students and families are participating in open enrollment and exiting their home/neighborhood schools. Next, I specifically describe how both the logistic based and survey based research was undertaken. 72 Logistic regression method. The logistic regression model or logit model used in this study allows me to predict which independent variables are most likely to be correlated with the family’s likelihood to participate in open enrollment, the dependent variable. The dependent variable measures a family’s participation in open enrollment (OE). OE is equal to one if the student is not attending their neighborhood school as a result of parent request, and zero if they are attending their neighborhood school. Using the data file obtained from GUSD, I compared the student’s neighborhood school with the school of actual attendance to establish the dichotomous variable of zero or one. This study uses multiple regression because there are more than one independent or predictor variable that is expected to influence the choice of leaving a neighborhood school. Regression analysis is appropriate because it allows for the calculation of the influence of one explanatory variable holding all other included explanatory variables constant. Dummy variables were created to represent differences in a student’s gender, home language, race and ethnicity, socio-economic status, and if the overall academic performance of the student’s home school (as measured by API1) was lower than that of the chosen school for the 2012-2013 school year. This logistic regression of the The API is a single number, ranging from a low of 200 to a high of 1000, which reflects a school’s, an LEA’s, or a student group’s performance level, based on the results of statewide testing. Its purpose is to measure the academic performance and growth of schools. The API was established by the PSAA, a landmark state law passed in 1999 that created a new academic accountability system for kindergarten through grade twelve public education in California. 1 The API is calculated by converting a student’s performance on statewide assessments across multiple content areas into points on the API scale. These points are then averaged across all students and all tests. The result is the API. An API is calculated for schools, LEAs, and for each student group with 11 or more valid scores at a school or an LEA. 73 dichotomous dependent variable, participation in open enrollment, yielded regression coefficients that indicate how a one-unit change in one of these predictors, holding other predictors constant, is expected to change the likelihood of a family’s participation in open enrollment. Throughout this study, I only discuss the influence of an explanatory variable if the confidence is greater than 90 percent that is exerts a non-zero influence on participation in open enrollment. Logistic regression analysis also provided a way to assess how well the predictor variables, overall, predict the likelihood of participation in open enrollment (Pallant, 2007). Similar to the t-test in linear regression, the Wald statistic used in binominal regression is used to assess the statistical significance of coefficients. Additionally, I interpreted the logit coefficient in the form of an “odds ratio.” Using this, I can predict by how much each variable increases the likelihood that a family will participate in open enrollment (Pampel, 2000). The logistic regression methodology relied on demographic and descriptive data of the 42,228 students enrolled in grades K-12 to determine which characteristics increased the likelihood that a family would participate in open enrollment. The area of human behavior, specifically related to the theoretical frameworks germane to this study, is inherently complex and therefore difficult to make accurate predictions. However, through a multiple regression I am able to identify those variables which together provide an estimation of a family’s/student’s likelihood of participating in open enrollment (Brace, Snelgar, & Kemp, 2012). Most importantly for the policy implications I am looking to draw, I can separate out the influence of one specific explanatory variable 74 holding the other constant. For instance, I can state with a specific degree of statistical confidence that if two students are identical in age, grade, home language, ethnicity, gender, and English proficiency, what effect low socio-economic status (or any of the other possible explanatory variables just mentioned) has on the decision of their parents for them to leave their neighborhood school. Survey-based method. The second analysis is a Pearson correlation and comparison of means coming from a second data source. It uses the results from parent surveys to analyze the nine variables described in Figure 3.5. Surveys were mailed to the study participants selected through the stratified random sampling procedures described earlier. The survey instrument used in this study was created by the researcher based on themes and issues uncovered during the review of relevant literature. The purpose of these two methods used together is more useful in fully answering my research questions. 75 Table 3.4. Descriptions of Variables in Survey Variable Description Participation in Open enrollment Participation is yes if the child is enrolled in a school other than the child's school of residence, also referred to home school or neighborhood school. The enrollment was a result of parents' requesting transfer through open enrollment and has not been done through involuntary transfer or other means. School Safety Measures the perception that the choice school is safer and freer from violence than the neighborhood school Principal leadership Measures the perception that the leadership at the choice school is more effective and desirable than the principal leadership at the neighborhood school School programs Measures the degree to which the choice school has more preferred academic and/or extracurricular programs than the neighborhood Measures the degree to which the choice was based on the greater perceived or real academic performance of the choice school over the neighborhood school School's academic performance Racial and ethnic diversity of school Measures the degree to which racial and ethnic diversity was a factor in choice. Distance and or transportation Measures the degree to which distance and/or transportation were factors in choice. Peer group influences Measures the degree to which parents believe their child will have an advantage by sending him or her to a school with higher achieving students. Note: These variables relate to both participants and non-participants in open enrollment. Parents were asked to comment on their neighborhood school, choice school, or both. The final and preliminary survey was piloted on six parents of elementary aged students who currently participate in open enrollment. These parents are known 76 personally to me and based on their knowledge of my study they were asked to pilot the survey. They were asked to reflect on time to complete survey, clarity of questions and wording, think-time for each question, redundancies, and any irritation or embarrassment any questions caused (Iarossi, 2006). The original survey included 41 questions. The test participants unanimously said it was too long and that several questions seemed to ask the same thing over again. When it was explained that the goal was to ask two similar questions in order to increase reliability, the test group agreed that the questions related specifically to school safety were unambiguous and only needed a few questions. Regarding clarity, questions related to principal leadership were described as “confusing” or “jargony.” The last relevant observation of the pilot group was that the pace of their response slowed when they reached the question asking if race diversity influenced their decision to stay at or leave their home school. During the debrief, I probed further on this observation and one respondent said, “I wasn’t sure if it was a good thing or a bad thing that I thought about race. It wasn’t my first consideration, but subconsciously I guess I was wondering about it (Anonymous, 2013).” As a result of the feedback from the pilot group, the survey was shortened to the present 18 questions and terms used specifically in education were replaced with more common terms. 77 Sampling. I used a proportional stratified random sampling of students from three selected high schools within the GUSD. The selection of these three schools was based on the results of the first logistic regression analysis and enrollment patterns for the 2012-2013 school year. When controlling for all other variables, parents at the three chosen high schools were likely to exercise choice. Chapter 4 will provide greater detail of the statistical analysis, however, the graph in Figure 3.4 shows the total numbers of 1) students that left each school, 2) students that chose the school in lieu of their neighborhood school, and 3) the total number of students that attend because it is their neighborhood school. Figure 3.4. Enrollment Patterns at Study High Schools The students were divided into two groups; leavers and stayers. Within each group of leavers and stayers, they were further divided by grade, 9-12. Ten percent from each of the 8 groups were randomly selected resulting in 487 participants. This method allowed for each to have an equal chance of selection and for each group to be equally represented in the study. The goal was to select a large enough representative sample so 78 that the results could be generalized back to the population and school district. This large sample size was selected in order to minimize the chance of sampling error. Although specific subgroup data is not studied, chapter five will outline future areas of study. The surveys sent to the 490 randomly selected parents and families uncovered those, social, economic, and perceptual factors that have led parents to exercise school choice other than their neighborhood school. Further, it guided me in understanding parents’ decision making and the degree of rationality in making schooling choices. The sample consisted of 490 participants; approximately 54% were female (n = 263), while the rest were male (n = 227). Chapter 4 will provide further descriptive characteristics of the sample. Correlation coefficients were computed among the nine school choice variables. The Pearson correlation coefficient index ranges in value from -1 to +1. A p value of less and .10 (or a 90 percent confidence of a non-zero influence) was required for significance. Returned surveys were assigned the dummy variable of one if the respondent was among the group who participates in open enrollment, or a zero if the respondent did not participate. Responses the Likert scale questions (1-14 on the survey) were recorded in SPSS to find the mean and standard deviation values for the independent variables. See Appendices B and C for complete surveys that were sent to both the families choosing to exercise choice and leave their neighborhood school, and those families choosing to stay at their neighborhood school. All forms of statistical analysis were conducted using Statistical Package for the Social Sciences (SPSS). This is a widely used tool for analyzing social science data. The 79 final level of analysis is the presentation of descriptive data comparing enrollment rates and trends of the three selected high schools and comparisons of racial and ethnic characteristics of the sites compared to state and GUSD averages. The goal is for the obtained results to be generalized across similar population. The combination of these two methodologies provides for a richer collection of variables that are necessary for policy recommendations and allows for generalization of the results across other similar sized and comprised schools and LEAs. These two analyses will be used together to improve the overall strength of the study (Creswell, 2009). Role of the Researcher I am the primary data collector in this study. I am a California high school administrator and previous middle school principal within the school district study site. It should be noted that I had no physical interaction with the participants, only with school district personnel for the purpose of collecting data and delivering surveys to be mailed to randomly selected participants. Protection of Participants All study related materials, surveys, consent forms, and related questions received University, participating school district, and dissertation committee approval prior to implementation. As a means of protecting participant confidentiality, respondents are not identified by name. I, assisted by the school district’s Department of Assessment, Evaluation and planning, was given access to student / parent information, and assisted 80 with distribution of survey instruments. I did not have access to or know the individual identities of the study participants. The names of the schools and school district were changed and pseudonyms’ for each school were used throughout this study. 81 Chapter 4 Data Analysis and Findings Introduction The purpose of this empirical portion of my study is to uncover the factors leading to the phenomenon of parents leaving their neighborhood schools and use that information to inform open enrollment and school improvement policies. Additionally, this empirical examination seeks to uncover the extent to which schools have the capacity to change or influence enrollment patterns of their schools; and to what degree do current school policies influence open enrollment and school choice decisions of parents. Within the frameworks of rational choice theory (RCT) and social cognitive theory (SCT), I investigate the following questions: 1. Through logistic regression analysis, what variables (available in this school district’s standard data files) exert a significant impact on the choice to participate in open enrollment, both in terms of them having a non-zero influence and the magnitude of this influence? 2. Through correlation and mean comparison, what factors (available through a selfadministered survey of parents in this school district) are significant in relation to why families choose to participate in open enrollment and leave their home/neighborhood school? 3. Through analysis of written survey responses, to what degree do current school policies influence open enrollment and school choice decisions? 82 School District Data Descriptive characteristics. In order to determine which of the demographic variables influence the likelihood of participation in open enrollment, I first used the standard data collected for every student in this district to perform several analyses. The goal in all of these analyses was gain a better understanding of the influence of a change in an expected causal variable to whether a parent chose to move their child from their neighborhood school to a different school within the same school district. Two different types of this analysis are presented next. First, descriptive statistics of demographic and socio-economic characteristics of all students, and dividing all into participants in open enrollment and non-participants, are presented. I then offer an independent samples t-test of whether the difference in the mean value of these characteristics is statistically significant between the two groups. Second, I present the results of the binary logistic regression using participation in open enrolment set equal to one as the dependent variable, and the demographic and socio-economic characteristics as explanatory variables. The sample size of 12,352 students represents all students in grades 9-12 and excludes those students enrolled in continuation school or any of GUSD special education centers. Table 4.1 offers the mean and standard deviation for the dependent variable and each explanatory variable for the entire population, and for both participants and non-participants in open enrollment. The table includes the minimum value, maximum value, mean, and standard deviation. 83 An independent-samples t test was conducted to evaluate the hypothesis that students with certain demographic characteristics exercise participation in open enrollment to a greater degree than students that do not possess those characteristics. This test evaluates the difference between the means of two independent groups: participants in open enrollment and non-participants. The t test evaluates whether the mean value of the test variables for one group, participants in open enrollment, differs significantly from the mean value of the test variables for the second group, nonparticipants in open enrollment (Green & Salkind, 2008). Table 4.1 reports the results of the difference in means test. It indicates, using an asterisk, that eight of these means are different in a statistically significant manner. In particular, pen enrollment participants were more likely to be assigned to a neighborhood school with an API lower than the district average, be female, or identify their race/ethnicity as other Asian. While non-participants were more likely to speak a home language other than English, come from a low socioeconomic status, identify their race and Latino, identify their race as South-East Asian. Figures 4.1 through 4.8 display the comparisons of group means that are statistically significant with a confidence level of 90 percent or higher, the standard deviation for the total population, participants, and nonparticipants, the F-ration to determine whether the variances in two independent samples are equal, the t-statistic, two tailed significance, the difference in mean, and standard error difference. 84 Table 4.1. Comparisons of Means of Student Characteristics Between Participants and Non-Participants Total Population of 12,354 (6,345 Participate in Choice, 6,009 Do Not) Mean Variable Name API of home school lower than district avg. Home language is other than English (All Races) Standard Deviation Total NonTotal NonParticipants Participants Population participants Population participants 0.388 0.481* 0.290 0.487 0.478 0.454 0.166 0.134 0.201* 0.500 0.341 0.400 Low SES 0.386 0.353 0.420* 0.372 0.478 0.494 Female .0531 0.566* 0.493 0.487 0.496 0.500 Home language is other than English (White) 0.062 0.051 0.074* 0.499 0.219 0.262 Hispanic or Latino 0.168 0.159 0.177* 0.241 0.366 0.382 African American 0.071 0.069 0.073 0.374 0.253 0.260 Native American 0.019 0.019 0.019 0.256 0.137 0.135 Pacific Islander 0.009 0.009 0.010 0.136 0.094 0.098 South-East Asian 0.009 0.002 0.005* 0.096 0.113 0.069 East Indian 0.003 0.002 0.004 0.113 0.049 0.062 Filipino 0.016 0.015 0.017 0.058 0.121 0.129 Korean 0.003 0.003 0.003 0.125 0.053 0.059 Japanese 0.001 0.001 0.001 0.056 0.033 0.026 Chinese 0.003 0.003 0.004 0.030 0.050 0.064 Other Asian 0.005 0.043* 0.035 0.058 0.203 0.183 Note: Participant n = 6345, Non-participant n = 6009, df = 12352. ^ Reported here is the difference between mean of this variable for non-participant less participant. * Indicates that a 90 percent level of confidence, the mean reported for this group is greater than the mean for the other group. 85 Figure 4.1. Comparisons of API Score of O.E. Participants and Non-Participants Figure 4.2. Comparisons of Home Language of O.E. Participants and Non-Participants Figure 4.3. Comparisons of Socioeconomic Status of O.E. Participants and NonParticipants Figure 4.4. Comparisons of Female O.E. Participants and Non-Participants 86 Figure 4.5. Comparisons of White, Non-English Home Language O.E. Participants and Non-Participants Figure 4.6. Comparisons of Hispanic or Latino O.E. Participants and Non-Participants Figure 4.7. Comparisons of South-East Asian O.E. Participants and Non-Participants Figure 4.8. Comparisons of Other Asian O.E. Participants and Non-Participants 87 Logistic regression. The comparison of means test only examines the difference in means of participants versus non-participants on one variable collected by the district. It does not, however, measure the influence of one of these characteristics on participation in open enrollment holding other explanatory variables constant. Therefore, a logistic regression analysis was conducted to determine the impact of one independent explanatory variable on the decision to participate in open enrollment after simultaneously controlling for all of the other explanatory variables included in the regression. This methodology can be thought of in this way. If two students and their parent(s) were identical in all of the ways race/ethnic and socio-economic status are measured here, except for one of these characteristics, the logistic regression coefficient will enable an estimation of the influence of this difference in one characteristic on the decision to participate in open enrollment. Note the desirable difference in this methodology from the previous comparison of means, which did nothing to control for other characteristics. The logistic regression model used sixteen independent variables that were included as likely causal factors, based in part on my earlier literature review, to influence participation. More could not be included because of the use of standard data collected by the school district. As shown in Table 4.2, seven of the independent variables were determined to exert an influence on participation that we can be confident is different than zero at a 90 percent or greater confidence level in a two-tailed test (p < .10). The Exp(B) column in Table 4.3 contains the natural exponent value of the respective logistic regression coefficient B. To determine how a one-unit change in a 88 respective explanatory variable influence the likelihood of participation, the appropriate calculation is (Exp(B)-1)*100 (Pampel, 2000). The value this returns indicates the percentage increase in likelihood that a student with this characteristic, holding other characteristics constant, is more or less likely to participate in open enrollment. Conversely, negative values indicate that a student with that characteristic is less likely to participate in open enrollment, holding other characteristics constant. Table 4.2. Logistic Regression Results B S.E. Wald Sig. Exp(B) API of home school lower than district 1.014 .041 610.717 .000 2.756* average Home language is other than English -.633 .078 72.746 .000 .515* Low socioeconomic status -.389 .042 84.363 .000 .678* Female .263 .037 49.672 .000 1.301* Home language is other than English .209 .110 3.586 .058 1.232* (White) Hispanic or Latino -.001 .062 .000 .992 .999 African American -.197 .076 6.690 .010 .821* -.044 .102 .750 .957 Native American .138 -.193 .956 .328 .825 Pacific Islander .197 -.137 .083 .774 .872 South East Asian .478 -.116 .581 Filipino .152 .446 .890 .582 1.504 Korean .474 .220 1.789 .594 .641 Japanese .742 .424 1.811 .201 .179 Chinese .475 .672 1.223 .729 4.266 Other Asian .353 .039 2.074* Note: *Indicates that this respective influence should be considered statistically significant from zero at a greater than 90 percent degree of confidence in a two-tailed test. See Table 4.3 for statistically significant regression coefficients and their corresponding variable that impacts open enrollment decisions. The variables in this table are ordered by the absolute value of their expected influence on a parents’ 89 participation in open enrollment. Also note that all of the variables are included in the model, not only the significant ones. The data suggests that the API of the student’s neighborhood school is the greatest indicator of whether that student will participate in open enrollment. Specifically, if the API of the neighborhood school is below the district average, then controlling for other variables, the student is 175.6 percent more likely to participate in open enrollment than a student whose neighborhood school’s API is above the district’s average. This provides strong evidence that parents leave or fail to enroll their children in neighborhood schools with academically underperforming students within them. Next in magnitude of influence are Other Asians who are 107.4 percent more likely to participate in open enrollment than the excluded category of race/ethnicity being Englishonly speaking white students. Among all races that speak a home language other than English, however, are 48.5 percent less likely to participate in open enrollment than this excluded category. Furthermore, African American students, or students of low socioeconomic status are 17.9 percent and 32.2 percent less likely to participate in open enrollment respectively. Finally, with all the student and family characteristics included in the logistic regression remaining constant, a female student is 30 percent more likely to participate in open enrollment than a male student, and white students whose home language is other than English are 23.2 percent more likely to participate in open enrollment. 90 Table 4.3 Statistically Significant Effects on Participation in Open Enrollment Effective of Variable Sig. Exp(B) Change in Effect on Participation in Open Variable Enrollment If API of neighborhood school is API of home school lower than the district average, the lower than district .000 2.756 175.6 student is 176% more likely to average* leave. Other Asian* Home language is other than English (All Races) .039 2.074 .000 .515 Low socioeconomic .000 status 0.678 Female 1.301 .000 Students who identify as Asian ethnicity (other than other Asian 107.4 categories are 107% more likely to leave than English-Speaking Whites (excluded category). Among all races, if student speaks a language at home other than -48.5 English, he or she is 48.5% less likely to leave. Low SES students are 32% less -32.2 likely to leave. 30.1 Female students are 30% more likely to leave. If student speaks a language at home other than English and is .058 1.232 23.2 White, he or she is 23.2% more likely to leave. African American students are 17.9% less likely to leave than African American* .000 0.821 -17.9 English-Speaking Whites (excluded category). Note: * Identifies those statistically significant variables that predict students will leave their neighborhood schools. Home language is other than English (White) The classification table in Table 4.4 uses the logistic regression results, and the specific characteristics of a student, to make a fitted prediction of the dependent variable (zero for no participation, 1 for participation) for them. If the fitted value is (less) greater than 0.5, they are predicted to (not) participate. The prediction is the compared to the 91 actual observed participation or not. The results of this comparison suggests that if we knew nothing about our variables and guessed that a person would participate in open enrollment that we would be overall correct 60 percent of the time. While the specific prediction for not participating in open enrollment would be correct about 57 percent of the time, and for participating would be correct about 63 percent of the time. Table 4.4. Fit of the Logistic Regression Prediction Classification Table a Predicted Participate in Open Enrollment 0 1 Observed Step 1 Participate in Open Enrollment Percentage Correct 0 3399 2610 56.6 1 2334 4011 63.2 60.0 Overall Percentage a. The cut value is .500. Analysis of Survey Results Descriptive statistics. As described in Chapter 3, surveys were mailed to 490 families selected through the stratified random sampling of students in grades 9 -12 at the three selected high schools that were identified in the logistic regression as having students with the greatest likelihood of participating in open enrollment. Six surveys were returned undeliverable. The total surveys returned for both participants and non-participants were 59 for a response rate of 12.1 percent. Demographic data was not asked on the surveys therefore it is unknown if the response were representative of the district as a whole. The goal of 92 the surveys was to collect and analyze unique parent perception data that is not available in any standard district files or the GUSD Intradistrict Transfer Application, see Appendix D for application. The survey was sent to parents of open enrollment participating students, and non-participating students. Part one of the survey asked five Likert scale questions related to school safety. Part two asked nine Likert scale questions related to principal’s leadership, academic, and social factors. Part three asked four yes or no questions related to school programs. See Table 4.8 for a descriptive analysis of each question, and Appendices A & B for the complete surveys. The scale for the surveys was as follows: 1. Not at all 2. A little 3. Moderately 4. Quite a bit 5. Extremely For questions 15 through 18 where the degrees of opinion were not needed, parents were asked a series of yes or no questions. Yes was set to a value of one and No was set to a value of zero. Table 4.5 Descriptive Statistics of Survey Responses (Questions 1-5) n Mean^ Standard Deviation Which of the following, if any, Total NonTotal NonTotal Nonare reasons why you left (or Sample participants Sample participants Sample participants Participants Participants Participants stay at) your neighborhood school? Suspensions and or expulsions 59 34 25 1.257 1.265 1.393 0.657 0.666 0.692 Illegal drug and alcohol use 59 34 25 1.497 1.394 2.329 1.007 0.814 1.675 Fights and conflict 59 34 25 1.528 1.455 1.955 0.977 0.891 1.338 Bullying, intimidation, and or 59 34 25 1.322 1.273 1.915 0.794 0.750 1.256 harassment Nothing specific but an overall impression that my neighborhood 59 34 25 1.823 1.818 1.374 1.175 1.192 0.857 school is unsafe Note: ^Variable range in value from 1 to 5; 1) Not at all, 2) A little, 3) Moderately, 4) Quite a bit, 5) Extremely. 93 Table 4.6 Descriptive Statistics of Survey Responses (Questions 6-14) Which of the following, if any, are n Mean^ reasons why you were drawn to (or Total NonTotal Nonstay at) your current / neighborhood Sample Participants participants Sample Participants participants school? The principal has a reputation of 59 34 25 2.089 2.092 2.584 success and accomplishment. The principal ensures that the needs of 59 34 25 2.242 2.249 2.835 all students are met. Nothing specific, but an overall impression that my home school is 59 34 25 2.457 2.412 2.996 "better administered" than any other school that I may have chosen. Results on statewide tests 59 34 25 3.032 3.033 2.916 Availability of honors or accelerated 59 34 25 3.000 3.000 3.376 classes Racial or ethnic diversity 59 34 25 1.654 1.673 1.960 Peer group influences 59 34 25 2.566 2.612 3.087 Nothing specific, but an overall impression that the learning 59 34 25 3.371 3.353 3.433 environment is better at my chosen school. Child-care or ease of transporting 59 34 25 2.265 2.302 3.080 student to school Standard Deviation Total NonSample Participants participants 1.269 1.288 1.288 1.329 1.348 1.312 1.421 1.417 1.472 1.424 1.446 1.320 1.495 1.518 1.379 0.988 1.413 0.997 1.407 1.274 1.498 1.592 1.612 1.438 1.650 1.660 1.824 Note: ^Variable range in value from 1 to 5; 1) Not at all, 2) A little, 3) Moderately, 4) Quite a bit, 5) Extremely. 94 Table 4.7 Descriptive Statistics of Survey Responses (Questions 15-18) n Mean^ Standard Deviation Please mark Yes or No to the following statements if they influenced your decision to 1) stay at your neighborhood Total NonTotal NonTotal NonParticipants Participants Participants Sample participants Sample participants Sample participants school, or 2) enroll your child in a school other than your neighborhood school. Preferred academic programs 59 34 25 0.383 0.392 0.750 0.486 0.392 0.433 Preferred visual and performing 59 34 25 0.125 0.183 0.417 0.322 0.183 0.493 arts programs Negative characteristics of 59 34 25 0.254 0.288 0.290 0.420 0.288 0.455 neighborhood school Positive characteristics of other 59 34 25 0.969 0.954 0.791 0.169 0.954 0.406 schools Note: Variable range in value of 0 or 1; 0) No, 1) Yes 95 96 Test for equality of means. Similar to the test conducted on the means of demographic data, an independentsamples t test was conducted to evaluate the hypothesis that certain parent perceptions work to influence the degree to which students participate in open enrollment to a greater degree than students whose parent(s) do not have those perceptions. Again, this test evaluates the difference between the means of two independent groups, participants in open enrollment, and non-participants. The t test evaluates whether the mean value of the test variables for one group, participants in open enrollment, differs significantly from the mean value of the test variables for the second group, non-participants in open enrollment (Green & Salkind, 2008). Tables 4.8 and 4.9 report the results of the difference in means test. They indicate seven of these means are different at a confidence level of 90 percent or greater. An additional two variables approach this threshold. For the dummy variables representing participants in open enrollment, only one, a positive characteristic of another school was higher for non-participants (0.968) as compared to participants (0.791.) As reported by the survey respondents, this suggests that the overall positive perceptions of a school besides their neighborhood school were the greatest indicator that a parent would choose to participate in open enrollment. Table 4.6 also shows, at a confidence of 89.9 percent, that the principal’s ability to meet the needs of all students results in a greater likelihood that parents remain at their neighborhood school than it draws them to other schools. While this may be a result of parents reporting their lived experiences and relationships with administration it reveals 97 that principal leadership can be a quantifiable characteristic. Parents of students that have never attended the neighborhood school have a limited ability to report on the competency of the school principal however. Also of note is the perception among participating parents that their neighborhood school is unsafe. Although parents did not indicate specifically that drugs, fights, or bullying were concerns, the difference in mean of 0.444 favored those participating parents. The responses tend to show that while parents could not point to one thing in particular as a concern, the overall perception that the neighborhood school was unsafe was greater for them than for non-participating parents. Childcare or ease of transportation was greater for non-participating families. This finding may be interpreted to suggest that distance from other schools may be a significant barrier for parents’ participation in open enrollment. Additional factors favoring non-participating parents were the availability of preferred academic programs and the availability of preferred visual and performing arts programs. With 75 percent of non-participating parents indicating that preferred academic programs was a factor in them staying at their neighborhood school, versus 36.5 percent of participating parents indicating that preferred academic programs is what drew them to another school supports the conclusion parents do exercise of degree of rationality when making school choice decision. Although the difference in mean was slightly less for preferred visual and performing arts programs (-.288), it was still a greater factor for non-participating parents. 98 Lastly, the most significant variable leading to participation in open enrollment was the overall perception of positive characteristics of other schools as compared to the neighborhood school. Nearly 97 percent of participant respondents indicated that the overall positive characteristics of other schools and not negative characteristics of the neighborhood school influenced their decision to participate in open enrollment. Table 4.8 Comparisons of Means of Survey Responses (Questions 1-14) Total Respondents 59 (34 Participate in Choice, 25 Do Not) Mean Standard Deviation Variable Name NonNonParticipants Participants Suspensions and or expulsions 1.265 participants 1.393 Illegal drugs & alcohols 1.394 2.329* 0.814 Fights and conflict 1.455 1.955* Bullying, intimidation, and or harassment 1.273 Nothing specific, but an overall impression that neighborhood school is unsafe 0.666 Sig. participants 0.692 .430 t Sig. (2Mean Std. Error tailed) Difference^ Difference -.718 .476 -.128 .178 1.675 .000 -2.834 .006 -.934 .330 0.891 1.338 .023 -1.722 .090 -.500 .290 1.915* 0.750 1.256 .007 -2.450 .017 -.642 .262 1.818 1.374 1.192 0.857 .050 1.584 .119 .444 .280 The principal has a reputation of success and accomplishment 2.092 2.584 1.288 1.288 .931 -1.452 .152 -.493 .339 The principal ensures that the needs of all students are met 2.249 2.835* 1.348 1.312 .606 -1.668 .101 -.586 .351 Nothing specific, but an overall impression of better administration 2.412 2.996 1.417 1.472 .768 -1.540 .129 -.584 .379 Results on statewide tests 3.033 2.916 1.446 1.320 .304 .317 .752 .117 .367 Availability of honors or accelerated classes 3.000 3.376 1.518 1.379 .569 -.976 .333 -.376 .385 Racial or ethnic diversity 1.673 1.960 0.997 1.274 .091 -.970 .336 -.287 .296 Peer group influences 2.612 3.087 1.407 1.498 .651 -1.247 .218 -.475 .381 99 Nothing specific, but an overall impression that the learning environment is better 3.353 3.433 1.612 1.438 .134 -.197 .845 -.080 .406 Child Care or Ease of Transportation 2.302 3.080* 1.660 1.824 .266 -1.705 .094 -.778 .456 Note: Participant n = 34, Non-participant n = 25, df = 57. ^ Reported here is the difference between mean of this variable for non-participant less participant. *Indicates that a 90 percent level of confidence, the mean reported for this group is greater than the mean for the other group. Variables range in value from 1 to 5; 1) Not at all, 2) A little, 3) Moderately, 4) Quite a bit, 5) Extremely Table 4.9 Comparisons of Means of Survey Responses (Questions 15-18) Total Respondents 59 (34 Participate in Choice, 25 Do Not) on 4 Yes or No Questions Mean Standard Deviation Variable Name Sig. t Sig. (2Mean Std. Error NonNonParticipants Participants tailed) Difference^ Difference participants participants Preferred Academic 0.365 0.750* 0.481 0.433 .071 -3.163 .003 -.385 .122 Programs Preferred Visual and 0.417* 0.326 0.493 .000 -2.704 .009 -.288 .107 Performing Arts Programs 0.129 Negative Characteristics of Neighborhood School Positive Characteristics of Other Schools 0.232 0.290 0.406 0.455 .239 -.516 .608 -.058 .113 0.968* 0.791 0.171 0.406 .000 2.278 .027 .176 .077 100 Note: Participant n = 34, Non-participant n = 25, df = 57. ^ Reported here is the difference between mean of this variable for non-participant less participant. * Indicates that a 90 percent level of confidence, the mean reported for this group is greater than the mean for the other group. Variable range in value of 0 or 1; 0 No, 1 Yes. 101 Summary Demographic characteristics including race, ethnicity, home language, and socioeconomic status are not statistically significant predictors of participation in open enrollment after controlling for those characteristics and using White English only speaking as the excluded category. As reported by survey respondents, among both participants and non-participants in open enrollment, the availability of honors or accelerated classes and the impression that the overall learning environment is better, led both groups to either stay at their neighborhood school or leave their neighborhood school to participate in open enrollment. While the surveys were strongly suggestive of trends and parent perception, the logistic regression results are the strongest evidence of who is participating in district open enrollment. Further, the number of surveys returned compared to the sample size, and the lack of any second language surveys returned at all limit the degree to which the survey results can be generalized across the entire district. Chapter 5 provides recommendations and additional ways to further this study. 102 Chapter 5 Conclusions and Recommendations Introduction The inspiration of this study came during my time as the principal of a middle school that experienced ten years of declining enrollment. At an enrollment peak of 1,200 students the school offered a wide variety of both academically rigorous courses and an ample selection of enrichment and elective classes. During my tenure as principal, we had 600 students and struggled to maintain a master schedule that could meet the needs of our growing diverse student population and be identified as a destination school in the community. As the school district developed and implemented its open enrollment policy, my school’s resources dwindled as the student enrollment shrank. Going through the school artifacts and speaking to those with historical knowledge, it became clear that the school had once been a gem, and with declining enrollment and other factors, the school had lost its luster. My first-hand experience witnessing students who chose to remain at their neighborhood school not have the same access to programs that their peers in other schools led to my alarm. I began to ask myself how I can influence school policy in so that no matter a child’s zip code, cultural background, or socioeconomic status, he or she could be assured that by remaining at their neighborhood school, they will have the same high quality education and programs offered to them. From that, I chose to use this 103 dissertation as a way to investigate this issue for those students and families that have historically been marginalized and forgotten about in our schools and communities. The remainder of this final chapter includes an overview of the purpose of the study and reintroduces the groups that made up the study participants. This chapter continues with a refamiliarization with the research methods and questions, a summary of the findings, and closes with five recommendations for school leaders intending to implement or improve open enrollment policy. Overview The purpose of this study was to reveal the factors leading to the phenomenon of parents leaving their neighborhood schools and use that information to inform open enrollment and school improvement policies. Additionally, this study sought to uncover the extent to which schools have the capacity to change or influence the enrollment patterns of their schools; and to what degree do current school policies influence open enrollment and school choice decisions of parents? The groups of parents analyzed in this study were both the parents of high school students who had chosen to enroll their child in a regular public school other than their assigned school of attendance, and those who remained at their neighborhood school. Research Questions This study uses both qualitative and quantitative methods research. I used these strategies to explore the characteristics of those who choose to participate, and not 104 participate in open enrollment, and to study the first hand perceptions of parents at the center of the phenomenon of enrolling their children in public schools other than their neighborhood schools. In order to understand the perceptions of parents choosing to enroll their child in a school other than their home school, this study specifically perform these research tasks: 1. What variables (available in standard data files) exert a significant impact on the choice to participate in open enrollment, both in terms of them having a non-zero influence and the magnitude of this influence? 2. Through correlation and mean comparison, what factors are significant in relation to why families choose to participate in open enrollment and leave their home/neighborhood school? 3. To what degree do current school policies influence open enrollment and school choice decisions? Summary of Findings The findings of this study suggest that the statistically significant indicators of participation in open enrollment are if the child’s neighborhood school has a lower API than the district average, if the child is female, students are identified as White but speak a language other than English at home, or if the child identifies as Other Asian when selfreporting race and ethnicity. The study also suggests that among all races of students whose home language is other than English, is of low socioeconomic status, or is African 105 American, they have a decreased likelihood of participating in open enrollment. Referring back to Table 4.3 in chapter four, you will recall that, if the API of the neighborhood school is below the district average, then controlling for other variables, the student is 175.6 percent more likely to participate in open enrollment than a student whose API of neighborhood school is above the district average. Next, are Other Asians who are 107.4 percent more likely to participate in open enrollment than the excluded category of race/ethnicity, which is English speaking, White. African American students are 17.9 percent less likely than this excluded category. Furthermore, a student of any race/ethnicity whose home language is not English is 48.5 percent less likely to participate in open enrollment. In contrast to this previous group, White students that speak a language other than English at home are 23.2 percent more likely to participate in open enrollment than the excluded category. A low socio-economic student is also 32.2 percent less likely to participate than a higher socio-economic student. Finally, holding the characteristics included in the logistic regression constant, a female student is 30 percent more likely to participate in open enrollment than a male. This study further finds that the overall positive characteristics of other schools was the most statistically significant perceptual factor leading to participation in open enrollment, while ease of transportation and the belief that the principal meets the needs of all students were the greatest perceptual factors in determining if parents stay at their neighborhood school. Preferred academic programs and visual/performing arts programs, specifically the International Baccalaureate, special education, band, drama, 106 and sports were factors for both participants and non-participants. However, the difference in mean between these two variables favored those who chose to stay at their neighborhood school. Research question 1. What variables (available in standard data files) exert a significant impact on the choice to participate in open enrollment, both in terms of them having a non-zero influence and the magnitude of this influence? Research question one was addressed using the quantitative data obtained from the school district. Through logistic regression analysis, Figures 5.1 and 5.2 display which characteristics were significant indicators (with a confidence level of 90 percent or greater) that students with these characteristics would have an increased or decreased likelihood of participating in open enrollment. Figure 5.1. Characteristics likely to increase and Decrease Participation in Open Enrollment 107 Figure 5.2. Graphical Representation of Numerically Significant Factors that Influence Open Enrollment Decisions – From Logistic Regression Research question 2. Through correlation and mean comparison, what factors are significant in relation to why families choose to participate in open enrollment and leave their home/neighborhood school? Research question two was addressed using the quantitative data from the parent surveys. As described earlier, participation in open enrollment was more closely tied to overall school academic performance and the overall positive characteristics of the school. Specific factors and perceptions linked to school safety were not closely related to participation in open enrollment. However, the rating of the overall impression that the neighborhood school was unsafe was slightly higher for participants than non- 108 participants. As reported in Chapter 4, the confidence of this finding at less than 90 percent suggests more that the lived experiences of those that stayed at their neighborhood school are stronger than the perceptions of families that have not sent their child to the neighborhood school. The response from participant parent in Case 26 reinforces the conclusion that lived experiences are more powerful than perceptions alone. “Older siblings went to neighborhood school and left due to environment there” (Anonymous Case 26, 2013). Figure 5.4 shows the results for the comparison of means from the parent surveys. Among the 14 Likert scale questions with ranges from 1 to 5, the parent perceptions with a numerically significant difference are listed in Figure 5.3. This figure further illustrates the degree to which non-participating parents rated transportation and school safety perceptions higher than parents participating in open enrollment. The parent survey also included four yes or no questions related to academic programs. Figure 5.4 illustrates that preferred academic and visual and performing arts programs were significant factors in parents choosing to stay at their neighborhood school. While overall positive characteristics of other schools was a significant factor in parents choosing to participate in open enrollment. The lack of district provided home to school transportation is an additional factor that forces many parents to keep their children in their neighborhood schools. Transportation and / or child care were greater factors for non-particpants in open enrollment than participants. 109 Figure 5.3. Mean Scores of Statistically Significant Parent Perceptions Regarding Academic Programs Research question 3. To what degree do current school policies influence open enrollment and school choice decisions? Research question 3 was addressed through the qualitative written responses to the parent surveys, the responses to questions 14 – 16 on the survey, and through an examiniation of the open enrollment application found in Appendix C. The findings suggests that programtic decisions and policies related to which schools offer which programs appear to be the significant policy related factors affecting school choice decisions. Additionally, transportation and distance appear to be factors that contribute to both participation and non-participation in open enrollment. Although transportation and child care were were more significant for nonparticipants, the following comments from particpating parents indicate that district boundaries force parents force parents to particpte if they want their child to attend the school closer to their homes. 110 “We live one mile from our current school. Our neighborhood school is six miles away” (Annonyous Case 9 participant, 2013). “None of these were the reason I didn’t send my children to the neighborhood school. The reason I didn’t send my children to the neighborhood school was because [it] was not the closest school to my house. I live closer to open enrollment schools than neighborhood schools” (Annonymouse Case 18 participant, 2013). “This school is closer and more academic” (Annonymous Case 10 participant, 2013). Figure 5.4 above shows that in regards to programs, three quarters of nonparticpating parents indicated that prefferred academic programs and nearly half of all non-participating parents indicated that preferred visual and performing arts programs were reasons why they stay at their neighborhood school. Table 5.1 list the programs that non-participating parents specifically identifed as preferred. 111 Table 5.1. Programs that Influence Parents’ Decision to Stay at Their Neighborhood School Programs That Influenced Parents to Stay at their Neighborhood School Advance Placement (AP) classes Art AVID (Advancement Via Individual Determination) Band Choir Drama Evening of the Arts GATE program High Achiever classes Home Economics International baccalaureate International studies Junior ROTC Morning sing PTA sponsored plays Talent shows Table 5.2 below, in contrast, lists the peferred programs specifically identified by particpating parents. Since not all parens submitted written responses, these two tables are not meant to represent all of the possible programitc choices. However, they are further indicators that parents do in fact exercise a certain degreee of rationality when making enrollmen decisions. 112 Table 5.2. Programs that Influenced Parents’ Decision to not Enroll Their Child at the Neighborhood School Programs That Influenced Parents’ Decision to not Enroll at their Neighborhood School Clubs Cross country team Drama Field trips GATE program International baccalaureate Japanese Marching band Desire to not be on a block schedule Polytechnic institute Radio and TV Rapid learner program Special education program Sports Track and field The following direct quotes from parents were provided on the paretn surveys. They provide greater insight into the thought process and specific reasons for particpation in open enrollment. “We stay at our neighborhood school because our student wants to stay with her friends. As parents, we prefer another high school” (Case 15, non-participant) “Band at home school needs more support monetarily and administratively” (Case 15, non-participant). 113 “My son is not involved with these programs, but I have seen on the district website programs offered at the school and some videos of what the students have done” (Case 11, non-participant). “We wish [School B], our neighborhood school, had a better music program” (Case 7, non-participant). In response to peer group influences, “It’s where all their friends are” (Case 7, nonparticipant). “The other important factor was that our current high school student wanted a different experience than his brothers. He wanted to go to a new school with new students and to make new friends” (Anonymous Case 1 participant, 2013) “Since transferring to [charter school], classes she was getting F in she is now getting a B” (Case 29, participant) In regards to questions 17 & 18, “Definitely both. If you take the Baccalaureate program out of [School B] test results, its one of the most poorly performing schools in the district. So if your student is not in that program it’s a bad environment. [School A] has outstanding test results across the board and college prep is encouraged for all students. But its ethnically/socioeconomically diverse as well” (Case 26, participant) 114 Discussion School choice advocates have long pointed to the benefits afforded to minorities and families in poverty that are able to exit their often failing school for another school. As discussed in Chapter 2, state and federal laws requiring that districts allow transfers of students that attend any one of the state’s consistently low-performing or persistently dangerous schools reinforce this belief. The statistically significant findings of this study reveal that when controlling only for race and ethnicity, African American students were less likely to participate in open enrollment than White students. Further, low socioeconomic students, students whose home language is other than English, and male students were also less likely to participate in open enrollment, see Figure 5.2. The findings of this study are consistent with those of Farrell, Jr. & Mathews (1990), George & Farrell, Jr., (1990), Jones-Wilson, Arnez, & Asbury (1992), and Johnson (2006). I contend that lack of transportation, cultural barriers, and economics are powerful forces in limiting the access of historically disadvantaged students to quality schools. Furthermore, transportation related issues are reasons for staying at the neighborhood school. Providing choice alone, without concerted efforts to change underperforming schools only avoids the complex and often difficult task of real school reform. The most importing factor, however, in choice to leave is if the neighborhood school has a lower API scoe than the district average. Returning to the theoretical frameworks that served as the backdrop for this study, it is clear that a certain degree of rationality and social awareness were factors in parents 115 excersicing choice for their child. Beginning with rational choice theory, we accept that human decision-making is based on their need to maximize self-interest. I can infer from the results of the parent surveys that motivaion, goal attainment, and personal advantage were driving factors in the decsion-making process for parents. The desire for preferred academic programs implies that parents undertook some research in examining programs. As described in Chapter 1, non-particpants in open enrollment are also choosers, in that they are making the choice to stay. Rational choice theory can also help to explain this. When transportation and child care are significant factors for non-particpants, it logically follows that it is irrational for those parents to change schools. Or it is more likely that keeping their child at their neighborhood school is a greater maximizer of their resources. Although not specifically answered by this study, it can be argued then that social cohesives, ability to particpate in after school activities, cost of transportation, and maintaining peer group are more or as important to parents as association with higher achieveing students. Social cognitve theory, in contrast, considers human emotions and experiences in examining decision-making. This study revealed that peer-group influences were not an overt driver of choice, however the fact that the most prevalant factor in choice was the home school’s API shows that parents did prefer to have their children attend school with higher performing students. I conclude that parents are more willing to describe their child as high achieveing and therefore want them to be around other similar ability students, rather than describe them as low-achieveing and want them to be around higher 116 achieving students. It is a nuance that is not explored in this study, but social cognitive theory supports the idea that self-identification, pereived social status, prestige, and social power are all powerful contextual factors related to choice. As described in Chapter 1, the core assumption regarding social cognitive theory is that environment, people, and behavior are three factors that are constantly influencing each other. Parents belived that by selecting the new school that were choosing the environment that provides the greatest opportunity for success and social support. The choice of a new school also provided the opportunity for observational learning with the goal of their child acquiring skills by watching the outcomes and actions of others. Lastly, by being surrounded by other high achieveing students, parents hoped for greater reinforcements through intrinsic rewards and incentives. SCT holds that “learners are proactively engaged in their own development and can make things happen by their actions” (Jabbarifar, 2011, p. 118). Parents of choosers then are relying on this theory to function in order for their child’s self efficacy to improve and they in turn increase their confidence in performing school related tasks. The findings of this study reveal that students whose home language is other than English, African American students, and lowa socioeconomic student are less likely to pursue those goals. Whereas females, other Asians, and non-English speaking White students are more likely to exercise school choice as a means of fullfilling the goals described above. 117 Policy Implications Although this study did not overtly seek to evaluate the educational effectiveness of choice programs, I did not find that choice policies alone improve the efficiency or effectiveness of traditional public schools. The forces of a market-driven system are extremely powerful when parents choose to leave underperforming schools. School principals and teachers cannot ignore the repelling affect of below average API scores. Equipped with this awareness, school district leaders must provide site principals and teachers with the needed resources in order to successfully transform into a high achieving school. This is true in and out of open enrollment environments. We also cannot rule out the potential benefits of such policies. Improved outcomes for choosers and matching families with programs appear to be the two most relevant outcomes of choice. Future and current policies must also consider equity and the impact of dissolving social cohesion. Further, this study confirms that the poor, African American, and students who speak a language other than English are not significant participants when it comes to school choice. Therefore, it is problematic for policy makers to predict that choice programs will more greatly benefit traditionally underserved students since there is yet to be strong or conclusive evidence in support of this. A discussion of policy would be incomplete without acknowledging the important role school boards, superintendents, principals, teachers, parents, and community play in the overall debate. Educating each on finance, equity, non-monetary resources, and 118 potential disparities in spending related to minorities is key (Darden & Cavendish, 2011). As California begins to rebound from the recession and more money is available to schools, school policy makers must resist trying to solve the education gap by adding programs alone. As a new weighted funding formula for schools makes its way through the state’s legislature, superintendents must ensure that all stakeholders are well informed and school choice policies remain only a part of a broader educational improvement strategy. Recommendations The following recommendations are intended for practitioners and officials at all levels of pre-K – 12 education. Support of open enrollment policies should be contingent upon a concurrent effort to support and improve underperforming schools. It is important to acknowledge that this broader discussion of school improvement would include decreasing drop out rates, the disproportionality of suspension and expulsions of Black and Latino students, chronic absenteeism, cigarette, alcohol, and drug abuse, teen pregnancy, school violence, teacher evaluation and retention, improved public health, and finance reform among others. Additionally, these recommendations are based on the current K-12 funding model and that no additional funding is available. Therefore, implementation of these recommendations may require the shifting of budget priorities. Based on the findings of this study, I recommend the following bold changes and or enhancement to any existing open enrollment policy: 119 1. Ensure that all high schools have the option to offer all of the programs that the community desires. 2. Provide more and better information to the most vulnerable groups regarding choices. (Speakers of languages other than English, African American, low socioeconomic) 3. Add transportation to foster more participation 4. Tap into the educational aspirations of the poor. 1. Ensure that all high schools have the option to offer all of the programs that the community desires. With the understanding that there are a finite number of marching-band leaders, teachers of languages other than English, music instructors, and others with specialty skills, school district leaders must ensure that those highly sought after programs are not clustered in geographic locations that favor any particular group. Additionally, schools should not be forced to decide between being using their teaching allotment to offer support classes or Advanced Placement classes. The results of the survey strongly indicate that programs, in additional to overall achievement, are powerful forces driving choice. It has already been establish that several disadvantaged groups are less likely to exercise choice. Rather than forcing them to go to certain programs, we have a moral imperative to bring those programs that we know increase academic success to them. 120 In the event the creation of new programs is cost prohibitive, the transfer of desirable programs to under-enrolled schools may be an option to dealing with the economies of scale. It is foreseeable that a policy of transferring programs in order to create desirable schools where they currently do not exist will lead to protest. Although I have spent little time discussing school leadership, this creates a tremendous opportunity for school leaders to build relationships across communities. The protest, however, may never come. Take for example the International Baccalaureate program. Based on the survey responses, it is the most desirable program of open enrollment participants and it is offered at only one site. That program then could be offered anywhere and still attract students. With high school enrollments in GUSD ranging from 700 to 2000 students, district leaders may consider capping enrollment in order to balance student body sizes. 2. Provide more and better information to the most vulnerable groups regarding choices. (Speakers of languages other than English, African American, lowsocioeconomic) A search of the GUSD high school web sites by me, someone who is well versed and knows well how to navigate K-12 education, discovered as many as six different names for what qualifies as a course catalog. They include 1) Course Catalog, 2) Course Description Handbook, 3) Majors and Career Exploration, 4) School Handbook, 5) Career Handbook, and 6) Course Handbook. This alone can confuse the savviest parents. 121 Therefore, providing clear, thorough, and accurate, and consistent information to parents must be a major tenet of any open enrollment plan. Parents who are unable to access the “grapevine” of knowledge described by Ball and Vincent (1998), or the adequate social capital, are at an extreme disadvantage compared to those parents that are well connected or have a network of other parents who provide information. Speakers of languages other than English and those with out access to these social networks are commonly left without the information they need which only further compounds the problem that school choice advocates are looking to solve. Providing parents with this information will lead to one of two outcomes. First, more information about the programs available will lead to more choice. Or second, parents will learn about the programs offered at their neighborhood schools and want to stay. The goal I am suggesting is to not force families out of their neighborhood schools, but to provide them with information in order to make informed decisions. Schools and school districts must endeavor to balance the informal information gathering and exchanges among parents with accurate, formalized means of disseminating information. Below list the main sources of information for parents: “Grapevine” of knowledge Parent information nights held by the school School website School information sheets provided at the central enrollment office School visitation days 122 Parents then are presumed to have accessed these resources, thoughtfully considered the pros and cons carefully weighing all options, and then come to an informed rational decision before being asked to complete a one-page transfer application. An emerging practice among school districts is to require parents who are requesting an out of district transfer to tour their home school. It is unclear if this practice is an effort to retain students at their home school or just provide an additional hurdle, but with 51 percent of GUSD students participating in open enrollment, extending that requirement to intradistrict transfers also would place too high a burden on school sites. What we do know from my research is that perception about a school and the actual lived experiences of parents that actually send their child to that school are not always congruent. Again, at the center of any school choice policy must be the commitment to providing all families with accurate and timely information. I further recommend the examination of the course catalogs for every high school to ensure parity. 3. Add transportation to foster more participation. School choice programs can sometimes further stratify and segregate communities. GUSD does not provide home to school transportation for students unless required by their individualized education plan. However, a plan that limits transportation to targeted groups low-income, low-performing, language minority can manage costs and still meet the educational needs of all of its students. Although 123 transportation in itself is expensive, the cost of remediation and credit recovery for high school students is even more so. I also recommend school districts map all schools and calculate drive times and distances from school to school and provide that information to parents along with transfer applications. 4. Tap into the educational aspirations of the poor. Status attainment was not specifically reviewed in this study, however the work of Peter Blau (1918-2002) and Otis Duncan (1921-2004) forced sociologists to ask the question of how do people attain status and under what conditions can they be upwardly mobile. Broadly, status attainment refers to a person’s position in society or class. Not only does family income and education influence attainment, so does one’s own ability and effort. Blau and Duncan (1967) discovered that men whose fathers held high-status jobs had increased odds of themselves attaining high status. They found a greater correlation, however, between education and status attainment. But what is clear from their research is that it was much better to start at the top than at the bottom (Lin, 1999). With this understanding coupled with the knowledge that not all families have the resources to participate in open enrollment, I recommend that GUSD form organized outreach to those communities least likely to participate in open enrollment and work to organize and invest their human resources to improve the educational outcomes of students in their communities. By leveraging the direct ties that each of them have, the 124 community will work together to advance the group as a whole rather than an individual student at a time. Figure 5.2 is simplified representation and modification of a social capital conceptual framework. It demonstrates how social cohesion and social capital work together to produce desired educational results. I recommend a combination of paid and volunteer workers to implement such a plan. The ability of community members to volunteer is directly tied to economic factors. As discussed in Chapter 1, higher rates of volunteering, among other things, result from higher education. Therefore, school leaders cannot make the mistake of presuming that communities in poverty will have the resources to sustain a completely volunteer run program. Figure 5.4. Conceptual Framework of Educational Goal Attainment 1. Connectedness 2. Values 1. 2. 3. 4. Social Support Social Leverage Informal Social Support Community Participation 125 While these recommendations advocate for expanded school choice, it would be irresponsible, unethical, and immoral to not consider the following arguments. With any of the recommendations it is important to account for the years of marginalization experienced by ethnic and language minority groups. Improvement efforts will often been met with suspicion, doubt, a sense of panic, or uncertainty (Ball & Gewirtz, 1997). If not done carefully, school choice programs can work to destabilize the very system that is the basis for our democracy. Widespread, unrestricted open enrollment may also drain the financial resources of the most needy communities and lead to further racial, ethnic, and economic stratification. The ultimate challenge for school districts will be implementing a policy that minimizes the harm to some while improving the outcomes for all. Future Research Under the umbrella a school choice are a variety of options. Included are interand intradistrict transfers, public charter schools, private schools, home schools, and online schooling. The results of this study, as it relates to intradistrict transfers may be enhanced by further research in the following areas: Qualitative Interviews While the parent surveys exposed many of the perceptions of participants and non-participants, one-on-one interviews or focus groups would further uncover the nuances associated with school choice. 126 Longitudinal Study of Participants in Open Enrollment The long-term benefits of open enrollment can only be known by matching identical cohorts of students from both groups and comparing their progress over time. Further Examination of Cultural Barriers to School Choice The knowledge that African American, low socioeconomic, speakers of languages other than English, and male students do not participate in open enrollment in significant numbers is not enough to inform policy. What is missing is insight into the kitchen table conversations that happen in homes across the state prior to and during the open enrollment period. Examination of different theories that may apply to school choice • This study briefly touched on goal attainment and social capital as additional forces driving parent choice. Additional research my further explore the link between these two theories and parent participation in open enrollment A new policy problem has emerged • The original policy problem statement asserted that choice leads to the deterioration of neighborhood schools in urban areas. What has emerged from my study is a second problem; increased choice has led to enrollment patterns that do not represent the communities of the neighborhood schools or choice schools. Future research may explore the link between the achievement gap, social participation, and enrollment patterns. 127 Appendix A Probability in Finding a School Attribute Important The change in the probability of finding a school attribute important, 1582 parents in New York City suburbs. Hi Scores Black Asian Hispanic Other College HS Grad Private New York .089* (.036) .078 (.070) .054 (.039) -.026 (.055) -.088* (.034) .155* (.035) .011 (.030) .025 (.031) .300 26.99 0.00 .03 Values -.097* (.025) -.057 (.048) -.115 (.031) -.041 (.043) .085* (.033) -.074** (.047) .163* (.030) -.029 (.028) .228 115.46 0.00 .13 Discipline Diversity .0633** -.001 (.031) (.016) 1.77* -.067* (.069) (.011) .177* - .063* (.032) (.018) .058 .036 (.050) (.035) -.066* .068* (.024) (.016) -.001 -.027 (.024) (.034) .040 -.085 (.021) (.010) .050* -.001 (.021) (.017) .135 .095 105.40 102.61 0.00 0.00 .14 .15 Obs. Prob. Chi2 P > Chi2 Pseudo R2 Notes: *probability <. 05, **probability <. 10, All significances are two tailed. Numbers n parentheses are standard errors. Numbers above standard error represent the change in probability for a discrete change in the independent variable from 0 to 1 (Schneider, Marschall, Teske, & Roch, 1998). 128 Appendix B Survey to Parents Participating in Open Enrollment Section One Which of the following were reasons why you left your neighborhood school? 1 Student suspensions and/or expulsions 2 Illegal drug and alcohol use 3 Fights and conflict 4 Bullying, intimidation, and/or harassment 5 Nothing specific, but an overall impression that my neighborhood school was "unsafe." not at all 1 a little 2 RATING moderately 3 quite a bit 4 extremely 5 not at all 1 a little 2 RATING moderately 3 quite a bit 4 extremely 5 Section Two Which of the following were reasons why you were drawn to your current school? 6 The principal has a reputation of success and accomplishment. 7 The principal ensures that the needs of all students are met. 8 Nothing specific, but an overall impression that my chosen school is "better administered" than my neighborhood school. 9 Results on statewide tests 10 Availability of honors or accelerated classes 11 Racial or ethnic diversity 12 Peer group influences 13 Nothing specific, but an overall impression that the learning environment is better at my chosen school than at my neighborhood school. 14 Child-care or ease of transporting student to chosen school Section Three Please mark Yes or No to the following statements if they influenced your decision to enroll your child in a school other than your neighborhood school. yes no yes no yes no yes no My neighborhood school did not offer the preferred academic programs. If yes, please list any programs 15 My neighborhood school did not offer the preferred visual and/or performing arts programs. If yes, please list any programs 16 Overall, the decision to leave my neighborhood school more based on negative characteristics that pushed me 17 away from it. 18 Overall, the decision to leave my neighborhood school was more based on the positive characteristics of my chosen school that pulled me toward it. 129 Appendix C Survey to Parents not Participating in Open Enrollment Section One Which of the following, if any, are concerns at your neighborhood school? 1 Student suspensions and/or expulsions 2 Illegal drug and alcohol use 3 Fights and conflict 4 Bullying, intimidation, and/or harassment 5 Nothing specific, but an overall feeling that my neighborhood school is "unsafe." not at all 1 a little 2 RATING moderately 3 quite a bit 4 extremely 5 not at all 1 a little 2 RATING moderately 3 quite a bit 4 extremely 5 Section Two Which of the following, if any, are reasons why you stay at your neighborhood school? 6 The principal has a reputation of success and accomplishment. 7 The principal ensures that the needs of all students are met. 8 Nothing specific, but an overall impression that my home school is "better administered" than any other school that I may have chosen. 9 Results on statewide tests 10 Availability of honors or accelerated classes 11 Racial or ethnic diversity 12 Peer group influences 13 Nothing specific, but an overall impression that the learning environment is better at my neighborhood school than any other school that I may have chosen. 14 Child-care or ease of transporting student to school Section Three Please mark Yes or No to the following statements if they influenced your decision to stay at your neighborhood school. 15 yes no yes no yes no yes no My neighborhood school offers the preferred academic programs. My neighborhood school offers the preferred visual and/or performing arts programs. If yes, please list any programs 16 Overall, the decision to stay at my neighborhood school is more based on negative characteristics of other 17 schools that keeps me here. Overall, the decision to stay at my neighborhood school is more based on positive characteristics of this school 18 that keeps me here. 130 Appendix D GUSD Intradistrict Transfer Application 131 References Abell, P. (1990). Denzin on rational choice theory. Rationality and Society , 2 (4), 495499. Adler, N. E., Boyce, T., Chesney, M. A., Cohen, S., Folkman, S., Kahn, R. L., et al. (1994, January). Socioeconomic status and health: The challenge of the gradient. American Psychologist , 49 (1), pp. 15-24. Akey, T., Plucker, J. A., Hansen, J. A., Michael, R., Branon, S., Fagan, R., et al. (2008). Study of the effectiveness and efficiency of charter schools in Indiana. Bloomington: Center for Evaluation and Education Policy. Anonymous. (2013, January 19). Survey Test. (C. A. Morris, Interviewer) Arsen, D., & Ni, Y. (2008). The competetive effect of school choice policies on performance in traditional public schools. East Lansing: Great Lakes Center for Educational Research & Practice. Aud, S., Fox, M. A., & Kewal-Ramani, A. (2010). Status and trends in the education of racial and ethnic groups. National Center for Education Statistics. Washington, DC: U.S. Department of Education. Ball, S. J., & Gewirtz, S. (1997). Grls in the education market: Choice, competition and complexity. Geneder and Education , 9 (2), 207-222. Ball, S. J., & Vincent, C. (1998). 'I heard it on the grapevine': 'Hot' knowledge and school choice. British Journal of Sociology of Education , 19 (3), 377-400. 132 Ballou, D., Goldring, E., & Liu, K. (2006). Magnet schools and student achievement. Vanderbilt University. New York: National Center for the Study of Privitazation in Education. Baranowski, T., Perry, C. L., & Parcel, G. S. (1997). How individuals, environments, and health behavior interact: Social cognitive theory. In K. Glanz, F. M. Lewis, & B. K. Rimer (Eds.), Health behavior and health education: theory, research, and practice (2nd Edition ed., pp. 153-178). Jossey-Bass. Beaudin, B. Q. (2003). Interdistrict Magnet Schools and Magnet Programs in Connecticut. Bureau of Evaluation and Educator Standards, Division of Evaluation and Research. Hartford: Connecticut State Department of Education. Bennett, M., & Sani, F. (2008). Childrens subjective identification with social groups: A self-stereotyping approach. Developmental Schience , 11 (1), 69-75. Betts, J. R., Rice, L. A., Zau, A. C., Tang, Y. E., & Koedel, C. R. (2006). Does school choice work? Effects on student integration and achievement. San Francisco: Public Policy Institute of California. Brace, N., Snelgar, R., & Kemp, R. (2012). SPSS for Psychologists (5th Edition ed.). New York: Palgrave Macmillan. California Department of Education. (2012, May). 2011–12 Academic Performance Index Reports: Information Guide . Retrieved June 10, 2012, from California Department of Education: http://www.cde.ca.gov/ta/ac/ap/documents/infoguide12.pdf 133 California School Boards Association. (2010, November). CSBA Sample Board Policy. Retrieved September 30, 2012, from California School Boards Association: http://www.csba.org/en/Services/Services/PolicyServices/~/media/938CE5EC32 DE4943A06530DD3903F05D.ashx Carlson, D., Lavery, L., & Witte, J. F. (2011). The determinants of interdistrict ope enrollment flows: Evidence from two states. University of Wisconsin-Madison, Educational Evaluation and Policy analysis. Madison: American Educational Research Association. Center for Education Reform. (2009). K-12 Facts. Retrieved 2011 йил 6-May from http://www.edreform.com/Fast_Facts/K12_Facts/ Christensen, B., Eaton, M., Garet, M. S., Miller, L. C., Hikawa, H., & DuBois, P. (2003). Evaulation of the magent schools assitance program, 1998 grantees. U.S. Department of Education, Office of the Under Secretary Policy and Program Studies Service. Washington, D.C.: American Institutes for Research. Cookson, P. W. (1994). School choice: The struggle for the soul of American education. Binghampton, NY: Vail-Ballou. Crain, R. L., Allen, A., Thaler, R., Sullivan, D., Zellman, G., Little, J. W., et al. (1999). The effects if academic career magnet education on high schools and their graduates. Berkeley: National Center for Research in Vocational Education. Creswell, J. W. (2009). Research design: Qualitative, quantitative, and mixed methods approaches (3rd Edition ed.). Los Angeles, CA: Sage Publications. 134 Culter, D. M., & Lleras-Muney, A. (2008). Education and health: Evaluating theories and evidence. (R. F. Schoeni, J. F. House, G. A. Kaplan, & H. Pollack, Eds.) Making Americans healthier: Social and economic policy as health policy , 29-60. Darden, E. C., & Cavendish, E. (2011). Achieving resource equity within a single school district: Erasing the opportunity gap by examining school board decisions. Education and Urban Society. Washington, DC: Sage. de Souza Briggs, X. (1997). Moving p versus moving out: Neighborhood effects in housing mobility programs. Housing Policy Debate , 8 (1), pp. 195-234. Dee, T. S. (1998). Competition and the Quality of Public Schools. Economics of Education Review , 17 (4), 418-427. Delpit, L. (1995). Other people's children: Cultural conflict in the classroom. New York, New York: New Press. Denzin, N. K. (1990). Reading rational choice theory. Rationality and Society , 2 (2), 172-189. Dillon, E. (2008). Plotting school choice: The challenges of crossing district lines. Washington, D.C.: Education Sector. Dynarski, S., Hoxby, C., Loveless, T., Schneider, M., Whitehurst, G., & Witte, J. (2010). charter schools: A report on rethinking the federal role in education. The Brookings Institute, Brown Center on Education Policy. Washington, DC: The Brookings Institute. 135 Education Commisson of the States. (2012). Education Commission of the States-Helping State Leaders Shape Education Policy. Retrieved July 25, 2012, from Choice of Schools: http://www.ecs.org/html/issue.asp?issueid=22&subissueid=326 Edwards, B., Crane, E., Barondess, H., & Perry, M. (2009). California's charter schools: 2009 update on issues and performance. EdSource. Mountain View, CA: EdSource. Farrell, Jr., W. C., & Mathews, J. E. (1990). School choice and the educational opportunities of African american children. The Journal of Negro Education , 59 (4), 526-537. Foster, K. M. (2004). Coming to terms: A discussion of John Ogbu's cultural-ecological theory of minority academic achievment. Intercultural Education , 15 (4), 369384. Fowler, F. C. (2002, September/October). The great school choice debate. The Clearing House , 76 (1), pp. 4-7. Frankenberg, E., Siegel-Hawley, G., & Wang, J. (2010). Choice without equity: Charter school segregation and the need for civil rights standards. Los Angeles, CA: The Civil Rights Project/Proyecto Derechos Civiles at UCLA. Frey, A., Ruchkin, V., Martin, A., & Schwab-Stone, M. (2009). Adolescents in transition: School and family characteristics in the development of violent behaviors entering high school. Child Psychiatry & Human Development , 40 (1), 1-13. 136 Fullan, M. (2006). Change theory: A force for school improvement. Centre for Strategic Education. Victoria: Centre for Strategic Education. Fuller, B., Burr, E., Huerta, L., Puryear, S., & Wexler, E. (1999). Abundant hopes, scarce evidence of results. University of California, Berkeley and Stanford University. Berkeley, CA: Policy Analysis for California Education. Gallup. (2011, August 19). Parents, Americans Much More Positive About Local Schools. Retrieved June 10, 2012, from Gallup: http://www.gallup.com/poll/149093/Parents-Americans-Positive-LocalSchools.aspx Gamoran, A. (1996). Student achievement in public magnet, public comprehensive, and private high schools. Educational Evaluation and Policy Analysis , 18 (1), 1-18. George, G. R., & Farrell, Jr., W. C. (1990). School choice and African American stuents: A legislative view. The Journal of Negro Education , 59 (4), 521-525. Goodman, R. H., & Zimerman, W. G. (2000). Thinking differently: Recommendations for 21st century school board/superintendent leadership, governance, and teamwork for high student achievment. New England School Development Council. Arlington: Educational Research Service. Green, S. B., & Salkind, N. J. (2008). Using SPSS for windows and Macintosh: analyzing and understanding data (5th Edition ed.). Upper Saddle river, NJ: Pearson. Green, S. (2002). Rational choice theory: An Overview. Baylor University. 137 Heebner, A. L. (1995). The impact of career magnet high schools: Experimental and qualitative. Journal of Vocational Educational Research , 20 (2), 22-55. Herrmann, A. M., Burroughs, N., & Plucker, J. A. (2009). Open Enrollment in K-12 Public Education. Center For Evaluation & Education Policy , 7 (3). Hill, P. T., & Lake, R. J. (2010). The charter school catch-22. Journal of School Choice: Research, Theory, and Reform , 4 (2), 232-235. Hoxby, C. M. (2009, May 5). Charter school research and economics. (S.-M. Institute, Interviewer) Hoxby, C. M. (2000). Does Competition among Public School Benefit Students and Taxpyers? The American Economic Review , 90 (5), 1209-1238. Hoxby, C. M. (2009, May 5). The promise and performance of charter schools: Drivers of educational improvement in the U.S.? Lecture . St. Louis, Missouri: St. Louis University. Hoxby, C. M. (2002). Would school choice change the teaching proession? The Journal of Human Resources , 37 (4), 846-91. Hoxby, C. M., & Murarka, S. (2009). Charter schools in New York city: who enrolls and how they affect their students' achievement. National Bureau of Economic Research. Cambridge, MA: National Bureau of Economic Research. Hsieh, C.-T., & Urquiola, M. (2006). The effects of generalized school choice on achievement and stratification: Evidence from Chile's voucher program. Journal of Public Economics , 1477-1503. 138 Iarossi, G. (2006). The power of survey design: a user's guide for managing surveys, interpreting results, and influencing respondents. Washington, DC: The World Bank. Jennings, J. L. (2010). School choice or schools' choice? Managing in an era of accountability. Sociology of Education , 83 (3), 227-247. Jimerson, L. (2002, September/October). Interdistrict open enrollment: The benign choice? The Clearing House , 76 (1), pp. 16-19. Johnson, H. B. (2006). The American dream and the power of wealth: Schools and inheriting inequality in the land of opportunity. New York, NY: Routledge. Johnson, H. B., & Shapiro, T. M. (2003). Good neighbors, good schools: Race and the "good choices" of white families. In A. Doane, & E. Bonilla-Silva (Eds.), White out: The continuing significance of racism. New York: Routledge. Jones-Wilson, F. C., Arnez, N. L., & Asbury, C. A. (1992). Why not public schools? Journal of Negro Education , 61 (2), 125-137. Kahlenberg, R. D. (2006). Helping children move from bad schools to good ones. New York: The Century Foundation. Koedel, C., Betts, J. R., Rice, L. A., & Zau, A. C. (2009). The integrating and segregating effects of school choice. Peabody Journal of Education , 84 (2), 110-129. Krueger, A. B., & Lindahl, M. (2000). Education for growth: Why and for who? Cambridge: National Bureau of Economic Research. 139 Lareau, A. (2002). Invisible inequality: Social class and childrearing in Black families and White families. American Sociological Review , 67 (5), 747-776. Lauen, D. L. (2007). Contextual explanations of school choice. Sociology of Education , 80 (3), 179-209. Ledwith, V. (2010, February). The influence of open enrollment on scholastic achievement among public school students in Los Angeles. American Journal of Education , 116 (2), pp. 243-262. Levin, H. M. (2012). Some economic guidelines for design of a chrter school district. Economics of Education Review (31), 331-343. Marshall, M. N. (1996). Sampling for qualitative research. Family Practice (13), 522525. Martinez, V. J., Goodwin, R. K., Kemerer, F., & Perna, L. (1995). The consequences of school choice: who leaves and who stays in the inner city. Social Science Quarterly , 76 (3), 485-501. McMahon, W. W. (2004). The social and external benefits of education. In G. Johnes, & J. Johnes (Eds.), International handbook of the economics of education (pp. 221259). Northampton, MA: Edward Elgar Publishing. McNamara Horvat, E., Weininger, E. B., & Lareau, A. (2003). From social ties to social capital: Class differences in the relations between schools and parent networks. american Educational Research Journal , 40 (2), 319-351. 140 Menard, S. (2002). Applied logistic regression analysis. University of Colorado, Institute of Behavioral Science. Thousand Oaks: Sage. Miron, G., Evergreen, S., & Urschel, J. (2008). The impact of school choice reforms on student achievement. Westen Michigan University, The Evaluation Center. Boulder, CO: Education Public Policy Interest Center. Morris, J. E. (1997). Voluntary desegregation in St. Louis, Missuri: Impact on partnerships among schools, African-American Families, and communities. St. Louis: Vanderbilt University. Moustakas, C. (1994). Phenomenological research methods. Los Angeles, CA: Sage Publications. Muhammad, A. (2009). Transforming school culture: How to over come staff division. Bloomington, IN: Solution Tree Press. National Household Education Surveys Program. (2010). Trends in the Use of School Choice: 1993 - 2010. US Department of Education. Ogbu, J. U., & Simons, H. D. (1998). Voluntary and Involuntary Minorities: A CulturalEcological Theory of School Perfomrance with Some Implications for Education. Anthropology & Education Quarterly , 29 (2), 155-188. Orfield, G., Losen, D., Wald, J., & Swanson, C. B. (2004). Losing our future: How minority youth are being left behind by the graduation rate crisis. The Civil Rights Project at Harvard University. Cambridge: Harvard Education Publishing Group. 141 Orfield, M., & Wallace, N. (2007). Expanding edcational opportunity through school and housing choice. CURA Reporter , 37 (2), 19-26. Pallant, J. (2007). SPSS Survival Maual: a step by step guide to data analysis using SPSS for Windows (3rd Edition ed.). New york: Open university Press. Pampel, F. C. (2000). Logistic Regression: A Primer. thousand Oaks, CA: Sage. Peterson, P. E., Woessmann, L., Hanushek, E. A., & Lastra-Anadon, C. X. (2011). Globally challenged: Are U.S. students ready to compete? Harvard university, Harvard's Program on Education Policy and Governance. Cambridge: Harvard University. Public Policy Forum. (2002, July 8). Districts say convenience for parents, not competition, is result of open enrollment. Research Bief , 90 (7), pp. 1-6. Public Policy Forum. (2012, February). Significant growth in school choice: More schools, more students, fewer limits on income eligibility. Reserch Brief , 100 (1), pp. 1-12. Ridenour, C., Lasley, II, T. J., & Bainbridge, W. L. (2001). The impact of emerging market-based public policy on urban schools and a democrataic society. Education and Urban Society , 34 (1), 66-83. Schneider, B., Swanson, C. B., & Riegle-Crumb, C. (1998). Opportunities for learning: Course sequences and positional advantages. Social Psychology of Education , 2, 25-53. 142 Schneider, M., Marschall, M., Teske, P., & Roch, C. (1998). School choice and culture wars n the classroom: What different parents seek from education. Social Science Quarterly , 79 (n), 489-501. Smrekar, C., & Goldring, E. (1999). School choice in urban America. New York, Neew York: Teachers College Press. Stuart, M. (1993). Choosing a secondary school: Can parents' behaviour be described as rational? Research. Stucke, M. E. (2009, December 14). Auditing self interest. America , pp. 10-14. Teske, P., & Schneider, M. (2001). What research can tell policy makers about school choice. Journa of Policy Analysis & Management , 20 (4), 609-631. The Civil Rights Project. (2010). Charter schools' political success is a civl rights failure. The University of California, Los Angeles. Los Angeles: The University of California. Thompson, M. N., & Dahling, J. J. (2012). Perceived social status and learning experiences in social cognitive career theory. Journal of Vocational Behavior (80), 351-361. Tooley, J. (1993). A Market-Led Aleternative for the curriculum: Breaking the Code. London: Tufnell. 143 U.S. Department of Education. (2007). Giving parents options: Strategies for informing parents and implementing public school school choice and supplemental educational services under No Child Left Behind. Washington, D.C.: Office of Innovation and Improvment. United States Census Bureau. (2012). Metropolitan and micropolitan. Retrieved June 6, 2012, from http://www.census.gov/population/metro/data/metrodef.html Wells, A. S. (1996). African-American students' view of school choice. In B. Fuller, R. F. Elmore, & G. Orfield (Eds.), Who chooses? Who loses?: Culture, institutions, and the unequal effects of school choice (pp. 25-49). New York, New York: Teachers College Press. Whitty, G., & Edwards, T. (1998). School Choice Policies in England and the United States: An Exploration of Their Origins and Significance. Comparative Education , 34 (2), 211-227. Young, T. W., & Clinchy, E. (1992). Choice in public education. New york, New York: Teachers college Press.