Educational policy analysis archives

Educational policy analysis archives

Material Information

Educational policy analysis archives
Arizona State University
University of South Florida
Place of Publication:
Tempe, Ariz
Tampa, Fla
Arizona State University
University of South Florida.
Publication Date:


Subjects / Keywords:
Education -- Research -- Periodicals ( lcsh )
non-fiction ( marcgt )
serial ( sobekcm )

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
E11-00463 ( USFLDC DOI )
e11.463 ( USFLDC Handle )

Postcard Information



This item has the following downloads:

Full Text
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam a22 u 4500
controlfield tag 008 c20059999azu 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E11-00463
0 245
Educational policy analysis archives.
n Vol. 13, no. 41 (October 08, 2005).
Tempe, Ariz. :
b Arizona State University ;
Tampa, Fla. :
University of South Florida.
c October 08, 2005
On school choice and test-based accountability / Damian W. Betebenner, Kenneth R. Howe [and] Samara S. Foster.
x Research
v Periodicals.
2 710
Arizona State University.
University of South Florida.
1 773
t Education Policy Analysis Archives (EPAA)
4 856

xml version 1.0 encoding UTF-8 standalone no
mods:mods xmlns:mods http:www.loc.govmodsv3 xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govmodsv3mods-3-1.xsd
mods:relatedItem type host
mods:identifier issn 1068-2341mods:part
mods:detail volume mods:number 13issue 41series Year mods:caption 20052005Month October10Day 88mods:originInfo mods:dateIssued iso8601 2005-10-08


Readers are free to copy display, and distribute this article, as long as the work is attributed to the author(s) and Education Policy Analysis Archives, it is distributed for noncommercial purposes only, and no alte ration or transformation is made in the work. More details of this Creative Commons license are available at http:/ / ses/by-nc-nd/2.5/. All other uses must be approved by the author(s) or EPAA EPAA is published jointly by the Colleges of Education at Arizona State University and the Universi ty of South Florida. Articles are indexed by H.W. Wilson & Co. Send commentary to Casey Cobb (c and errata notes to Sherman Dorn (epaa-editor@s EDUCATION POLICY ANALYSIS ARCHIVES A peer-reviewed scholarly journal Editor: Sherman Dorn College of Education University of South Florida Volume 13 Number 41 Octo ber 8, 2005 ISSN 1068–2341 On School Choice and Test-Based Accountability Damian W. Betebenner Boston College Kenneth R. Howe Samara S. Foster University of Colorado Citation: Betebenner, D. W., Howe, K. R., & Foster, S. S. (2005). On school choice and test-based accountability. Education Policy Analysis Archives, 13 (41). Retrieved [date] from Abstract Among the two most prom inent school reform measures currently being implemented in The United States are sch ool choice and test-b ased accountability. Until recently, the two policy initiatives remained relatively distinct from one another. With the passage of the No Child Left Be hind Act of 2001 (NCLB), a mutualism between choice and accountability emerged whereby school choice complements test-based accountability. In the first portion of this study we present a conceptual overview of sc hool choice and test-based ac countability and explicate connections between the two that are ex plicit in reform implementations like NCLB or implicit within the market-based reform literature in which school choice and test-based accountabilit y reside. In the second po rtion we scrutinize the connections, in particular, between school choice and test-based accountability using a large western school district with a popular choi ce system in place. Data from three sources ar e combined to explore the ways in which school choice and test-based accountability draw on each ot her: state assessment data of children in the district, school choice da ta for every participating stud ent in the dist rict choice


Education Policy Analysis Archives Vol. 13 No. 41 2 program, and a parental surve y of both participants and n on-participants of choice asking their attitudes concerning the use of school report card s in the district. Results suggest that choice is of benefit academically to only the lowest achieving students, choice part icipation is not uniform across different ethnic groups in the district, and parents’ primary motivations as reported on a surv ey for participation in choice are not due to te st scores, though this is not consistent with choice preferences among parents in the district. As such, our results generally confirm the hypotheses of choice critics more so than advocates. Keywords: school choice; acco untability; student testing. Introduction During the last two decades, The United States has witnessed a sweeping tide of reform efforts directed toward the improvement of public education. Largely initiated by the seminal report A Nation at Risk (National Commission on Excellen ce in Education [NCEE], 1983), the quest for improved education in the United States ve ntured in numerous directions—from the varied curriculum wars of the 1980s to the standards based reform efforts of the 1990s. With reauthorization of the Elementary and Secondary Education Act of 1965 (ESEA), referred to as the No Child Left Behind Act (NCLB), no abatement is apparent in the desire to reform education at the beginning of the 21st century. Two reform initia tives currently dominate public education policy: Test-based accountability and parental choice. Given their inclusion in NCLB, both are likely to pervade discussions concerning education reform fo r years to come and to have lasting effects on public education in The United States. NCLB is without precedent in both scope and direction, employing test-based accountability and school choice as the fundamental mechanisms to engender school improvement. The legislation implements this accountability-choice mutualism along two lines: First, NCLB encourages the expansion of school choice by allowing students attending Title I schools to transfer to another public school (including charter schools) if their school has been identified as being in need of school improvement, corrective action, or restructu ring as defined by the accountability components of the law. NCLB provides technical assistance for schools failing to make adequate progress and also requires school districts to provide free trans portation for students who choose to attend other district schools. Second, NCLB requires statewide a ssessments in grades 3–8 in reading and math, and science testing in at least one grade in el ementary, middle, and high school. Furthermore, beginning in the 2002–03 school-year, states were required to provide report cards for each of their public schools. In simplest terms, the rationale for combining choice and accountability is that the two complement one another: (1) accountability facilit ates parental choice by supplying parents with information (in the form of school report cards) th ey need to make informed choices, and (2) choice facilitates accountability by fostering competition for enrollment. With the choice/accountability nexus as the foundation, our intent in this study is twofold: First, we wish to analyze the theory that brings accountability and choice together under one federal policy. In particular we wish to unpack the rationale linking choice and accountability using the language of market-based reforms and make explicit the manner in which accountability and pa rental choice are thought to complement one another. Second, using data compiled from a la rge western school district, we ascertain whether market-based reforms that link accountability and choice function as prescribed. We begin with a brief review of market-based educatio n reforms vis--vis school choice.


On School Choice and Test-Based Accountability 3 Market-based School Reforms Milton Friedman (1962) is widely credited with originating the idea of reforming public education via school choice. He proposed a market -driven scheme in which public funding and administrative authority would be transferred to private schools where parents redeemed their government supplied vouchers. Various other vouc her proposals have followed that have added the overall improvement of public education to Friedm an’s more modest goal of increased efficiency (e.g., Chubb & Moe, 1990). A few limited voucher programs have been implemented, with Milwaukee’s “Parental Choice Program” being the most long-lived and carefully studied (see Witte, 2001). The Florida A+ program (a hybrid system, which includes both private and public schools) has gained considerable attention of late (Greene, 2001). Since vouchers were declared constitutional by the United States Supreme Court in 2002, a numb er of states are in the process of considering legislation that would open the door to vouchers in their states. Colorado, for example, recently passed a state voucher program that allocates funding to disadvantaged children from kindergarten through grade 12 whenever their local school distri ct demonstrates low academic performance. The program was recently struck down by the Colorado State Supreme Court. But proponents intend to reintroduce a voucher bill that will pass constitution al muster. It’s likely that other states will follow Colorado’s lead in using choice as a major initiative to reform public education. Not all proponents of public school choice embr ace a market rationale, but this rationale now sets the terms of the debate (Hess, 2002). Pr oponents argue that market regulation is far better than bureaucratic government regulation (Chubb & Moe, 1990). The market mechanism of parental choice allows funds for schools to follow students and creates a system of competition among schools. They argue that this competition provides sc hools with a mandate to im prove or risk losing money and even closure (Greene, Peterson, & Du 1997; Hess, Maranto, & Milliman, 1999; Hess & Leal, 2001; Merrifield, 2001). The result is an incentive for schools to provide better services (through, for example, innovation, specializati on, efficiency, etc.) and to increase student achievement. Other supporters contend that such programs can serve to promote equity (Coons & Sugarman, 1978). In general, those who see public choice as a means of promoting equity observe that parents have long chosen schools by choosin g their place of residence (Henig & Sugarman, 1999). Parents’ incomes and social positions thus largely determine their power to choose. A choice policy that removes attendance boundaries permits students to choose schools independent of the price of houses in the neighborhoods in which they liv e. It thus provides all parents with choice, and also promises to promote diversity in schools Critics of market-driven school choice question whether it can improve achievement overall. They contend that the market may simply redistri bute students as a result of “skimming,” where certain schools’ mean achiev ement increases only because ot her schools’ mean achievement decreases (e.g., Carnoy, 2000). Critics also contend that by introducing competition, cooperation that currently exists between teachers and schools will be sacrificed. The majority of criticism for marketbased reforms like school choice falls on the issue of equity, charging that such reforms are much more likely to exacerbate inequity than to mitiga te it. School choice plans, for example, typically make no provision to protect students from being harmed while schools are declining, before they are reconstituted (e.g., Arsen, Plank & Sykes, 2000 ; Lauder & Hughs, 1999). School choice can also result in stratification by ethnicity and income (e.g., Cobb & Glass, 1999; Howe, Eisenhart & Betebenner, 2001), exclude special needs students (e.g., Arsen et al., 2000; Rothstein, 1999; Zollers, 2000), and thereby force other public schools to carry the burden of accommodating the needs of more difficult to teach students. Finally, critics al so claim that there is inequality among parents in their capacity to choose because certain parents ma y lack the information needed to participate in


Education Policy Analysis Archives Vol. 13 No. 41 4 meaningful deliberation, and others may lack trust in authorities (Fuller & Elmore, 1996; Wells, 1993). In a comprehensive overview of the eff ects of school competition and educational outcomes, Belfield and Levin (2002) conducted a revi ew of research from the United States in order to determine “to what extent, and according to what measures of output, does increased competition improve educational quality” (2002, pp. 2 79–280). The criteria for the articles chosen were that they must address educational outcome s and competitive pressures across large markets and contain an analysis on the basis of largescale, cross-sectional data sets. Competition was constructed in terms of greater school choice fo r parents and students, and educational outcomes were mostly measured in terms of standardized test scores, as well as graduation/attainment, expenditures/efficiency, teacher quality, students’ post-school wages, and local housing prices. They identified about 40 relevant empirical studies and determined that the research “shows reasonably consistent evidence of a link between competition (choice) and educational quality” (p. 297). In other words, they found a positive correlation between increased competition and educational outcomes/quality, but the actual effects were “sub stantively modest” (p. 297). They raise caution in using these findings to support policies to promote greater competition among schools. They suggest that the benefits of increased competition must be set against any additional generated costs to justify specific policy approaches. They write, “The benefits of competition…should not be exaggerated. … a number of them may in fact be the ‘same’ benefit, but calculated in different ways” (p. 296). Additionally, they warn that equity issue s stemming from increased competition must be considered since “market systems rank poorly ag ainst equity criteria (e.g., by showing greater segregation and partitioning of student groups)” (Belfield & Levin, 2002, p. 296–297). School Choice and Test-based Accountability Despite their marriage in NCLB, school c hoice and test-based a ccountability were not originally envisioned as complementary policy initiatives. Their paths toward policy prominence have been different and often uneven. Though the roots of both reside in the 1960s, it was testbased accountability that first gained a foothold under Title I of the Elementary and Secondary Education Act of 1965. With its implementation more than three decades ago, large scale testing has been the primary means of evalua ting program efficacy and the deter mination of federal support for the education of low achieving students in poor neighborhoods (Heubert & Hauser, 1999). Since then, student testing for accountability purposes has only grown in prominence. The mid-1980s marked an expansion in the use of students’ standardized test results for accountability purposes (Linn, 2000, 2001; Nationa l Research Council (NRC), 1999). In 1986, 33 states required some form of minimum competenc y testing of its students (Office of Technology Assessment, 1992). Following the Charlottesville Summit in 1989, state and national leaders launched a more ambitious conception of accoun tability based upon standards-based reforms. The tests used for this new system of accountability were to be aligned to world-class standards and were intended to be more intellectually challenging than were those of the 1980s. As the 1990s unfolded, standards-based accountability became the “touchst one” for state governance, as states moved away from judging schools in terms of inputs (Elmor e, Ablemann, & Fuhrman, 1996) and focused on results. By the mid-1990s, for example, test based requirements for high school graduation existed in 18 states (Bond, Braskamp, & Roeber, 1996). Toward the latter half of the 1990s, a subtle yet important shift in test-based accountability occurred: the results of assessments began to be distributed as broadly as possible to the general public. The primary vehicle for providing this information has been “school report cards”. By 2001,


On School Choice and Test-Based Accountability 5 27 states had implemented some form of school report cards as part of their accountability systems (Orlofsky & Olson, 2001), and, as mentioned prev iously, report cards are mandated by NCLB. Test results are the primary ingredient of the report card s, which also include information such as safety, class demographics, graduation and dropout rates, student mobility, and information on student teacher ratios (Goertz, Duffy, & Carlson-LeFl och, 2000; Education Commission of the States, [ECS], 1999). Today, school report cards are ava ilable online, from the district, and are often published in local papers. The rationale underlying support for test-base d accountability—whether in the 1960s or currently—is that it enables school improvement by tying rewards and sanctions to measurable outcomes as defined by standardized test scores. Whereas school administrators have a great deal of power to alter schools based upon test results, parents, regardless of what information is available to them, have been largely powerless to implement sc hool level change. Thus, in order to extend the rationale underlying support for test-based accountability to situations where parents become the primary consumers of the testing results, parents n eed to have at their disposal some means by which to effect change upon the school. Giving pa rents the choice to determine what school their child attends is one such means. Some early studies (e.g., Wells, 1993; Wells & Crane, 1992) suggest that only savvy and powerful parents avail themselves of accountability in formation, and that they use it to promote only their own children’s interests. More recent studies (e.g., Schneider, Teske & Marshal, 2000) concur that only a relatively small group of activist pare nts use such information, but go onto suggest that the behavior of these parents (ev en if self-interested) produces overall improvement. To date, no studies have been conducted on the universal school report card requirements of the type required by NCLB—in which all parents receive accountability information, not just those who actively seek it out—to address fundamental questions such as: What kinds of parents use the information provided, for what purposes, and to what effects? The debate over public school choice is current ly driven by a collection of isolated and conflicting studies. In their recent comprehensive review of the existing empirical evidence, Gill, Timpane, Ross, and Brewer (2001) assert that no general conclusions are warranted with regard to the efficacy of school choice prog rams. The authors cite two major shortcomings with regard to existing research on school choice: (1) Current re search is of the “black box” variety—not delving deeply into how choice affects the actual operation of the school and merely focusing on inputs and outputs, and (2) current research has focused on prog rams that have been operating for a relatively short period of time with few participants. Anothe r shortcoming is that the conflicting studies have examined different educational policy contexts or, in the case of Cobb and Glass (1999) versus Hoxby (2001), used different data and analys is methods to examine the same context. The most contentious issue for school choice po licies historically—and that now extends to policies such as NCLB that integrate choice and ac countability—is their effect on patterns of school enrollments by race/ethnicity, income, and special education. Some scholars have argued on behalf of choice—for vouchers targeted at low-income stud ents, in particular—that it provides a means by which to increase racial integration (e.g., Greene, 1999), but this claim has been disputed (e.g., Reardon & Yun, 2002). The effects of choice on pa tterns of enrollment, it would seem, are highly dependent on the details of specific policies and the social and historical context in which they operate (e.g., Gorard, et al. 2001, Wells, 1998). Several studies of comprehensive choice programs (i.e., programs not confined to targeted vouchers ) have supported the more modest claim that while choice may not increase integration neither does it increase stratification (e.g., Gorard, et al.; Hoxby, 2001). This conclusion has been disputed, ho wever, by studies where choice has increased stratification (e.g., Carnoy, 2000; Cobb & Glass, 1999; Fiske & Ladd, 2000; Howe, et al. 2001).


Education Policy Analysis Archives Vol. 13 No. 41 6 Another family of studies of the effects of choice has focused on the academic achievement of students enrolled in choice schools (e.g., charter schools and means-tested vouchers). No general conclusions have emerged regarding whether c hoice schools outperform assigned public schools with comparable students. Moreover, these studies are limited because they fail to examine the effects of choice policy on students in the system remaining in assigned neighborhood schools. A comprehensive evaluation must also include th e effect of choice on achievement (positive or negative) of students remaining in assigned schools (e.g., Gill, et al. 2001; Hoxby, 2001) Several recent studies have adopted this more comprehensive approach. For example, Gorard, Fitz, and Taylor (2001), Greene (2001), and Hoxby (2001) have each produced empirical studies supporting the argument that competitive pressures created by choice policies (in the form of charter schools and/or vouchers) foster impr oved achievement overall, inclusive of low performing schools. Although thes e studies are an advance over the kinds of studies described previously, they have important limitations with re spect to the kinds of policies examined vis--vis NCLB-like policies. In particular, each was based on school or district level data (as opposed to student level data) with only the Gorard et al. study employing longitudinal data. The Current Study Following recent studies, the goal of this st udy is to examine school choice and test-based accountability together in an empirical fashion so as to provide a more comprehensive report on each and a more insightful synthesis of market-ba sed reforms in general. To this end our study addresses questions in three areas: School choice and student achievement An outcome suggested by much of the school choice literature is that students participating in a c hoice program will outperform their counterparts who do not. Our study investigates this question and also includes the effect of school choice on achievement of students not participating in choice and remaining in their neighborhood schools. School choice and patterns of student enrollment Another set of analyses in this study focus on the nature of the students and schools participat ing in school choice. Analyses here fall into two broad categories: (1) what are the characteristics of students participating in school choice, and (2) what are the characteristics of the schools th at are most desirable in terms of choice? Parents’ use of accountability information Data on achievement and the movements of students within a district allowing for choice is crucial to understanding how achievement and choice interact. But it is the dissemination of achievem ent data to parents that is at the heart of the school choice/test-based accountability mutualism. To include this component in our study we surveyed parents both participating and not participating in choice and asked them about their use of school report cards and other information with regard to beliefs about their children’s schools. Method The location for the present study is a larg e western school district with an approximate enrollment of 27,500 students. Historically, the qua lity of the district’s schools has ranged from very good to excellent. School choice has existed in th e district since 1961. Th e choice program in the district, called open enrollment, allows parents from both inside and outside the district to send their child to any school in the district in which there is space available after enrollment by neighborhood children. By the mid 1990s, spurred by parents who were unhappy with the district’s implementation of the “middle school philosophy” or who compla ined about a perceived lack of emphasis on


On School Choice and Test-Based Accountability 7 academics in the district more generally, various c hoice options began to proliferate. Coincident with these developments, a new school board sympathetic to choice wa s elected, and the superintendent responsible for the middle school philosophy resigned This was also a time when the school choice movement began to accelerate at both the state and national levels. As open enrollment expanded in the distri ct, four choice options were added to the traditional option of enrolling in any neighborhood school on a space-available basis: (1) focus schools which offer a particular curricular foc us; (2) neighborhood focus schools which offer the standard district curriculum; (3) strand schools whic h offer the standard district curriculum alongside a different curricular strand; and (4) charter schools, whose accountability to the district is specified by a contract. In 1999–2000, 21 of 57 district schools had incorpor ated one of the types of choice options just described.1. This compares with only five such schools providing such choice in 1994– 95, all emphasizing either bilingual or experien tial education. Thus, between 1994–95 and 1999– 2000, 16 additional articulated choice schools were added, half of which adopted the mission of an explicit emphasis on academic rigor and college preparation. Core Knowledge was most prominent among the new options provided with five schools adopting it. Currently, more than 25 percent of students now take advantage of open enrollment to attend the district schools other than those assigned to them by attendance area. Concurrent with the boom in school choice in th e district the state began a statewide testing program. Implemented in 1997, the program tested 4th grade students in reading and writing. The program has since expanded and currently tests students from 3rd to 10th grade in reading and writing, from 5th to 10th grade in mathematics, and science in the 8th grade. For the past three years, annual test results as well as numerous other data a ssociated with each school are tabulated by the state department of education and reported to the public as school accountability reports (SAR). These reports are available to the public from a number of different sources including local newspapers and both the district and state education websites. Whereas the effects of school choi ce are typically hard to isolate, the district considered here is a relatively closed system where schools must compete for enrollment from the same pool of students. Thus, the district provides an ideal setting to scrutinize the broad assertions leveled by proponents and skeptics of school choice in particul ar and market-based reforms in general. Even so, there are a number of confounding factors to be considered before any actual analysis is performed. The ideal “experimental” situation woul d provide for a pretest, a treatment consisting of the parents choosing (or not choosing) which school their child will attend, and post-tests to see what the results of choosing are. The two prim ary components of such an analysis are student assessment and parental choice data. Limitations in the availability and nature of such data require that only a small subset of all students in the district be used in the analysis of school choice and academic achievement. Based upon previous school choice research in the district (Howe et al., 2001), we identify two groups of parents participating in choice in the district: Parents whose children are exercising choice between school level and parents whose childr en are exercising choice within school level. Children participating in choice between school le vels are children who are enrolling in either kindergarten, 6th grade, or 9th grade and thus are, overall, transiti oning into a new elementary, middle, or high schools.2 Choice exercised within school level occurs when children switch elementary, middle or high schools in mid-stream. These two groups of parents indicated fundamentally 1 One of two K–8 schools, 11 of 33 elementary schools, 5 of 13 middle schools, and 4 of 9 high schools. 2 The schools in the district under study are pr imarily K–5 elementary, 6–8 middle, and 9–12 high schools. There are two sch ools designated as K–8.


Education Policy Analysis Archives Vol. 13 No. 41 8 different motivations for participating in the dist rict choice program. Within school level parents often exercised choice in the hopes of alleviating sp ecific difficulties that their child was having at their current school. Thus, the rationale for partic ipating in choice for these parents was highly idiosyncratic. By contrast, between school levels parents, particularly at the kindergarten and 6th grade levels, had more homogeneous concerns cen tered around their child attending the highest quality elementary or middle school available. In a ddition, the great majority of choice participants, 75 percent, were between school levels. Among the between school level parents exercisi ng choice, further limitations exist. There are numerous parents who enroll their child in kindergarten but those children are not subject to pre-testing and thus do not allow for any analysis to be performed concerning the results of choice. The ideal group of children for analysis is the group of children making the transition from elementary to middle school in the district: Acad emic measures exist which allow for a pre-post comparison, they are numerous, and their parents motivations for participating in choice (i.e., wanting their child to attend the best middle school possible) are relatively homogeneous. Unless otherwise indicated, the analyses performed here concern students matriculating from the 5th grade in 2000–01 to the 6th grade in 2001–02. Data for the study came from three sources: district assessment data, district choice records, and a pare nt survey of beliefs about school report cards: Assessment data Test data for all students in the district from the inception of the statewide testing program were provided by the district. Assessment data were pared to only include 5th and 6th grade students in the district during the 2000–01 and 2001–02 academic years, respectively. Tests in reading and mathematics were given to students in those grades and in those years. Choice data School choice records were provided by the district for analysis. These records include data regarding 3 ranked school choices as well as the neighborhood school the student would normally attend. After substantial data clea ning, these records were combined with the test data when a matching student record could be fo und. After matching, 1,961 cases remained for analysis. Survey data A telephone survey was administered in sp ring 2003 to elicit parental beliefs about the school report cards. Two populations of parents were identified: The population of parents sending their child to their neighborhood school (non-choosers) and the population of parents participating in the choice program. A st ratified random sample was constructed using location within district and whether the parent’s child was entering kindergarten, sixth, or ninth grade. The total number of chooser and non-c hooser respondents was 200 and 202 respectively. Because survey respondents were guaranteed anonymity, no linking of parent to student data was possible. The merged assessment and choice data was used to examine questions regarding students achievement as well as examining patterns of c hoice based enrollment. Survey data was used to assess parent attitudes towards the school report card s in the district. The methodology employed to address each of the three areas follows. School Choice and Student Achievement To quantify the academic value of choice with respect to student achievement, we employed a series of multilevel growth models (Goldstein, 20 03; Singer & Willett, 2003). In general, the model employs three levels: Occasion k (level 1) within Stud ent j (level 2) within School k (level 3). Within this three level structure a series of four nested models were analyzed: (1) an unconditional means model; (2) an unconditional growth model with no fixed effects; (3) an unconditional growth model taking account of whether or not the student is a participant in choice or not; and (4) an


On School Choice and Test-Based Accountability 9 unconditional growth model taking account of both student participation in choice and initial academic ability. Models were examined for both reading and mathematics. The unconditional means model, sometimes referred to as the empty model, partitions variance between the three levels of the model and is employed as a baseline with which to compare later more complicated models. The model is give n in composite form in Equation 1. The equation partitions the math/reading scale score into school ( 00k), student ( 0jk), and occasion ( ijk) components. Yijk = 000 + 00k + 0jk + ijk (1) The unconditional growth model fits a linear trajectory to each student in the data set. Because there are only two years of data currently av ailable, there will be no residual error at level one due to a perfect fit. Two unconditional growth models are of interest: A growth model which sets growth to differ between students but is fi xed between schools and a growth model allowing growth to vary between both students and schools. The purpose of examining the two models is to determine whether the more parsimonious model which fixes growth across schools is sufficient to model the data. Equation 2 presents the unconditional growth model allowing for slopes to vary across individuals but fixes growth across schools. Equation 3 allows growth to vary across both individuals and schools. The residuals 00k, 0jk, and ijk, are identical to those defined in the unconditional means model of Equation 1. The residuals 1jk and 1k are residuals for slope at levels 2 and 3, respectively Yijk = 000 + 100(GRADEijk 5) + 00k + 0jk + ijk + 1jk(GRADEijk 5) (2) Yijk = 000 + 100(GRADEijk 5) + 00k + 0jk + ijk + 1k (GRADEijk 5) + 1jk(GRADEijk 5) (3) To model the effects of choice on achievement, a fixed effect is added to the unconditional growth model. The fixed effect has three levels depending upon whether the student did not participate in school choice, participated in sc hool choice and accepted their enrollment at a school of their choosing, or participated in school c hoice but attended their assigned neighborhood school. Based upon the results of Model 2 versus Model 3, gr owth is allowed to vary across both individuals and schools. We consider two models in this context: A parsimonious model accounting for choice that sets all choice groupsÂ’ growth rates to be equal and a model that allows for growth rates to vary between choice groups. Equation 4 provides the equation used to estimate the choice effects while fixing growth rates between groups. Equation 5 is the equation allowing for different growth rates for the three choice groups. Yijk = 000 + 010CHOICEj + 100(GRADEijk 5) + 00k + 0jk + ijk + 1k(GRADEijk 5) + 1jk(GRADEijk 5) (4) Yijk = 000 + 010CHOICEj + 100(GRADEijk 5) + 110(GRADEijk 5)CHOICEj + 00k + 0jk + ijk + 1k(GRADEijk 5) + 1jk(GRADEijk 5) (5)


Education Policy Analysis Archives Vol. 13 No. 41 10 The final multilevel model investigating the effects of choice on academic achievement incorporates a further fixed effect indicating the quartile of the student on the initial measure. The purpose of this is to control for initial ability sin ce choosers are generally high er performing students in this district than non-choosers and regression toward the mean is likely to affect the attribution of growth rates to the choice groups. The fully cro ssed model allowing for initial status quartile and choice group is given in Equation 6. Yijk = 000 + 010CHOICEj + 020QUARTILEj + 021QUARTILEj CHOICEj + 100(GRADEijk 5) + 110(GRADEijk 5)CHOICEj + 120(GRADEijk 5)QUARTILEj + 121(GRADEijk 5)QUARTILEj CHOICEj + 021 + 00k + 0jk + ijk + 1k(GRADEijk 5) + 1jk(GRADEijk 5) (6) School Choice and Patterns of Student Enrollment Analyses on choice based student enrollment fall into two categories: First, we investigate what factors are most likely to pr edict whether a student is a participant in choice or not. Next, we investigate what are the characteristics of school s that draw people via the choice program. The purpose is to look at patterns of enrollment due to choice both in terms of the students and in terms of the schools. To analyze choice patterns associated with students we employ a two-level logistic regression model using chooser/non-chooser as th e dichotomous outcome variable (Snijders & Bosker, 1999). In this setting, level 1 represents the student and level 2 represents the 5th grade school the student was attending when they applied to be part of the choice program. One of the benefits of employing the two level design is that it allows for the modeling of between school variation which is known to exist (i.e., students ar e much more likely participate in choice in some elementary schools than in others). Like with the analysis of achievement and choice, a series of models is used to isolate the most prominent factors associated with choice participation. We begin with the unconditional means model (or empty model). This model, given in Equation 7, expresses the log-odds of the proba bility of being a chooser in school j, logit(Pj), as a linear function of the grand mean of all such probabilities, 0, plus random deviations from this average for each group, U0j logit(Pj) = 0 + U0j (7) Next, we wish to test whether the likelihood of participating in choice is dependant upon the academic performance of the student measured th e year prior to when the choice decision would take effect. When there are different likelihoods for high achievers versus low achievers, we refer to this as process as “skimming”. To test whether “skimming” is occurring with regard to test performance, 5th grade reading and math scores were added as covariates to Model 7. This model is given in Equation 8. logit(Pij) = 0 + 1READij + 2MATHij + 0j (8) Lastly, we extended the model of Equation 8 to include a fixed factor for ethnicity. This inclusion allows one to test whether or not “skimming” is occurring with regard to the ethnicity. Because there are so few African American and Nati ve American students in the district, we restrict


On School Choice and Test-Based Accountability 11 reporting, but not analysis, to Asian, Hispanic, and White students. This model is given by Equation 9. The estimates of the model indicate whether the likelihood of participating in choice is dependent upon the ethnicity of the student. logit(Pij) = 0 + 1READij + 2MATHij + 3ETHNICITYij + 0j (9) Results Results from an examination of school choi ce and student achievement were interesting. The baseline results for Model 1 are presented in Tabl e 1. Here again, level-one represents occasion, level-two represents students, and level three represents schools. As one would expect, most of the variability (66.4% in math and 66.9% in reading) in the observed scores is at the student level. At the school level in math and reading the amount of variability accounted for is 13.3% and 13.7%, respectively. Extending Model 1 to include a covariate for time (Models 2 and 3), not surprisingly, provided a much better fit for the data. Model 3, which allows for differing growth rates between both students and schools provided a significantly better fit than Model 2, which allowed for differing growth rates between students and fixed growth rates for schools. In particular, for Model 3, -2LogLikelihood for Math and Reading were 40,7 99.720 and 38,965.620, respectively, whereas the values derived for Model 2 (which estimates tw o fewer parameters) were 40,815.540 and 38,982.190, respectively. Table 1 Parameter estimates for Model 1, math and reading exams Math Reading Fixed Effects Coefficient S.E. Coefficient S.E. 000 547.0 6.4 645.9 5.1 Random Effects Var. Comp. S.E. Var. Comp. S.E. Level-three (school) effects var(00k) 703.9 248.9 451.1 159.1 Level-two (student) effects var(0jk) 3525.9 133.1 2197.9 82.3 Level-one (occasion) effects: var(ijk) 1078.5 35.2 633.5 20.7 -2LogLikelihood 41319.0 39353.6 Models 4 and 5 add a fixed effect for choice and differ with regard to their allowance for variable growth rates between three different c hoice groups (chooser who left (CL), choosers who stayed (CS), and non-choosers (NC)). The results are given in Table 2. The most important point to be derived form the results presented is that Model 5 was not significantly better at fitting the data than Model 5(the difference in -2LogLikelihood between the two models for math and reading were 1.44 and 0.05, respectively) indicating that allowing equal growth rates of the three choice groups is not a hypothesis that can be rejected. This is evid ent given the very slight change in the variance components between the models within each subject. That is, the addition of choice status does not


Education Policy Analysis Archives Vol. 13 No. 41 12 help account for any of the residual variance associated with the model at across the student or school level. Overall, the results of Models 4 and 5 together confirm that th ere is no benefit to the academic performance of students that can be associated with their choice status in the district in the first year. Table 2 Parameter estimates for Models 4 and 5 for math and reading exams Math Reading Model 4 Model 5 Model 4 Model 5 Fixed Effects Coeff. S.E. Coeff. S.E. Coeff. S.E. Coeff. S.E. 000 532.7 6.7532.7 6.7 635.1 5.2 635.0 5.2 100 22.3 1.722.5 1.9 14.7 1.4 14.9 1.5 010 CL 7.7 4.57.8 4.6 9.2 3.5 9.6 3.8 010 CS 9.5 9.812.5 10.1 10.7 7.7 11.6 8.4 110 CLxGrade — — -0.1 2.6 — — -0.5 2.2 110 CSxGrade — — -7.6 6.3 — — -1.4 5.0 Random Effects Var. Comp. S.E. Var. Comp. S.E. Var. Comp. S.E. Var. Comp. S.E. Level-3 school effects var( 00k) 733.4 259.5734.9 260.3 436.3 156.3 433.9 155.6 var( 1k) 34.5 17.335.6 17.8 22.2 11.0 22.4 11.1 cov( 00k, 1k) -88.4 52.7-89.9 53.5 -58. 133.3 -56.6 33.3 Level-2 student effects var( 0jk) 4273.8 138.94272.6 138.9 2963.1 96.4 2963.8 96.4 var( 1jk) 1623.8 53.31624.3 53.2 1011.2 33.2 1012.1 33.2 cov( 0jk, 1jk) -589.3 62.5-589.0 62.5 -656.0 42.9 -656.8 42.9 Level-1 occasion effects var( ijk) 0.0 0.00.0 0.0 0.0 0.0 0.0 0.0 -2LogLikelihood 40796.0 40794.6 38957.5 38957.5 Building on Model 4 and 5, the final model used to examine school choice and student achievement, Model 6, incorporates a fixed effect denoting the quartile of academic performance the student reached in the 5th grade, prior to their choice/non-choice. Including this term allows for an examination of whether choice help s/hinders students at some performance levels that are masked by the overall results of Model 3. Results of this model for both math and reading are presented in Figure 1.


On School Choice and Test-Based Accountability 13 2001-2002 MathScale Score Grade 5Grade 6 400450500550600650NC NC NC NC NC NC NC NC CL CL CL CL CL CL CL CL CS CS CS CS CS CS CS CS 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile2001-2002 ReadingScale Score Grade 5Grade 6 500550600650700750NC NC NC NC NC NC NC NC CL CL CL CL CL CL CL CL CS CS CS CS CS CS CS CS 1st Quartile 2nd Quartile 3rd Quartile 4th Quartile Figure 1: Grade 5 to grade 6 math and read ing growth estimates for non-choosers (NC), choosers leaving (CL) and choosers staying (CS) by quartile The “gold-standard” comparison is between th e choosers who left (CL) and the choosers who stayed (CS) since both groups are equivalent in the fact that they requested to leave their neighborhood schools. Because the number of CS stud ents is small (most students that participate in choice get some school option which they take), the differences in growth rates between the CS and CL groups, though large in some cases, are not statistically significant. Comparisons between non-choosers (NC) and CL yielded inter esting and statistically significant results. In particular, with regard to the math test, those students in th e 1st quartile who chose out of their neighborhood school demonstrated significantly better performance on the sixth grade math test than did their counterparts who stayed. In the 2nd and 4th quartiles this ordering was reversed: students choosing out of their neighborhood schools who left perf ormed significantly worse than did their counterparts who stayed. The results in reading demonstrated less difference between the three groups. There were no significant differences between the growth rates of the three groups in each of the four quartiles. Overall, the results suggest that no broad claims can be made about the efficacy of choice on student academic achievemen t as measured by the state assessment. Results of the examination of patterns of student enrollment with regard to choice status indicate “skimming” is occurring. The lo gistic regression model of Equation 8 yields coefficients that were significant and positive for reading but not significantly different than zero for math. Specifically, 1 = 0.00103 (0.00117) and 2 = 0.00517 (0.00146). Thus, the probability of participating in choice increase as the student’s fifth grade re ading score increases. In a choice system with no “skimming”, one would expect to see no relationshi p between student ability and their participation in choice. It is unclear why reading yielded a significant coefficient but not math. Certainly more data is necessary to draw long term conclusions abou t skimming, a likely result is movement towards higher concentrations of high ability students at certain schools and higher concentration of low ability students at other schools—tracking-at-large so to speak.


Education Policy Analysis Archives Vol. 13 No. 41 14 When looking at patterns of student enrollment including ethnicity as a factor some interesting results arise. Figure 2 presents the results of the analyses associated with Equation 9. Across all score levels, Asian students demonstrated a significantly higher probability of participating in choice than did Hispanic or White students. There was virtually no difference between White and Hispanic students with regard to their probability of participation. The results imply, at least in terms of those people choosing to participate in the choice system, that “skimming” based upon ethnicity is not occurring in the district given the relatively small number of Asian students in the sample. Just as the previous results present a complicated picture of what the consequences of choice in the district are, results from the survey administer ed to parents in the district also indicate that the motivations behind the parents’ decision to participate or not in school choice are nuanced. Specifically we found that parents, when asked about the importance of assessment based school ratings in the decision of where to send their child, did not value the school-level test results as highly as other factors including safety and school curriculum. Parents participating in school choice did value school report cards and the test based information significantly more than did those parents not participating in school choice, but fo r both groups, there were other non-testing factors that were reported to be more importa nt in their decision making process. 5th Grade Reading ScoreProbability of Choice Participation 500525550575600625650675700 Asian Hispanic White Figure 2: Probability of partic ipating in choice based upon 5th grade reading scor es and ethnicity This is partly consistent with results derived from district choice data showing what schools are most popular. When looking at middle schools, where test data is available and parental influence is greatest on the school the student at tends, the most requested middle schools for choice are those schools with the highest report card rati ngs. Thus, it appears based upon both survey data and patterns of choice in the district, that test scores do play a significant role in the decision making


On School Choice and Test-Based Accountability 15 process of parents. But, there are other considerati ons that trump test score results in the parents’ overall decision. Conclusion In this study we have examined arguments c oncerning school choice and its relation to testbased accountability in three ways: By examining how choice affects student achievement as measured by statewide achievement tests in math and reading, by examining how choice affects patterns of student enrollment across schools in the district as measured by statewide assessments, and by examining how choice is informed by results of statewide assessment that are reported in school report cards. With regard to the assertions put forward by the proponents and opponents of choice, taking the district we examined as a crucible of choice this paper lends greater support to the contentions of opponents of c hoice than the proponents: There was not a uniform benefit on student test scores for those students participating in choice. Thus, the contention that allowing choice will help academic achievement is not supported by our findings. In fact, the only students in our study who showed a positive benefit from choice were those students from the lowest quartile. But their increase was only demonstrated on the math test. It was not confirmed by their performance on the reading test. Our results suggest that “skimming” by ability is occurring in the district. This result may go some way to explaining why, when simple descript ive statistics are analyzed at, it appears that schools with a high amount of choice do well. Th ese schools may be equal in quality to other, less desirous schools in the district, but the pool of students from which they draw is highly able, leading some to wrongly believe that it is the school th at is responsible for these children’s scores. Parents’ decisions about what schools to send their children to are not uniformly directed by test scores alone. This is true for both parents participating in choice and those who don’t. There are many factors that influence parental attitudes a bout schools including reputation, safety, location, school where the child’s friends will attend, and curriculum. In future work we hope to extend the current anal yses to include more years of data that will shed light on what the longer term effects of choi ce are. The examination of choice and its impact upon achievement and patterns of student enrollmen t should be central to any discussion about the efficacy of choice in the public education system. It is our hope that similar examinations to ours will be carried out in others districts with school choice systems in place so that a more complete picture is available of this burgeoning phenom enon. In that way we all can better understand whether school choice does truly improve education in the United States. References Arsen, D., Plank, D., & Sykes, G. (2000). School choice policies in Michigan: The rules matter (Tech. Rep.). East Lansing: Michigan State University. Belfield, C. R. & Levin, H. M. (Summer 2002). The effects of competition between schools on educational outcomes: A review for the United States. Review of Educational Research, 72 (2), pp. 279–341.


Education Policy Analysis Archives Vol. 13 No. 41 16 Bond, L. A., Braskamp, D., & Roeber, E. D. (1996). The status of states student assessment programs in the United States (Tech. Rep.). Oak Brook, IL: North Central Regional Educational Laboratory and Council of Chief State School Officers. Chubb, J. E., & Moe, T. M. (1990). Politics, markets, an d America’s schools Washington, DC: The Brookings Institution. Cobb, C., & Glass, G. (1999). Ethnic se gregation in Arizona charter schools. Education Policy Analysis Archives, 7 (1). Retrieved November 14, 2003, from at Coons, J., & Sugarman, S. (1978). Education by choice: The case for family control Berkeley, CA: University of California Press. Education Commission of the States. (1999). Education accountability systems in the 50 states (ECS Publication No. SI–99–8W). Denver, CO: Education Commission of the States. Elmore, R., Ablemann, C., & Fuhr man, S. (1996). The new accountability in state education reform: From process to perf ormance. In H. Ladd (Ed.), Holding schools accountable: Performance-based reform in education (pp. 65–98). Washington DC: The Brooking Institute Press. Friedman, M. (1962). Capitalism and freedom Chicago: University of Chicago Press. Fuller, B., & Elmore, R. F. (Eds.). (1996). Who chooses? who loses? Cultur e, institutions, and the unequal effects of school choice New York: Teachers College Press. Gill, B., Timpane, M., Ross, K., & Brewer, K. (2001). Rhetoric versus reality: What we know and what we need to know about vouchers and charter schools (Tech. Rep.). Santa Monica, CA: RAND. Goertz, M., Duffy, M., & Carl son-LeFloch, K. (2000). Assessment and accountabi lity systems: 50 state profiles (CPRE Research Report Series No. RR–0 46). Philadelphia, PA: Consortium for Policy Research in Education. Retrieved Sept ember 18, 2003, from at Publications_Accountability.htm. Goldstein, H. (2003). Multilevel statistical models (3rd ed.). London: Arnold. Gorard, S., Fitz, J., & Taylor C. (2001). School choice impacts: What do we know. Educational Researcher, 30 (7), 18–23. Greene, J. P. (2001). An evaluation of the Florida A+ accountability and choice program (Tech. Rep.). New York: The Manhattan Instit ute. Retrieved June 4, 2003, from at http://www.manhattan-instit Greene, J. P., Peterson, P., & Du, J. (1997). Effectiveness of school choice: The Milwaukee experiment (Occasional Paper 97–01). Cambridg e, MA: Harvard Program on Education Policy and Governance.


On School Choice and Test-Based Accountability 17 Henig, J., & Sugarman, S. (1999). The nature and extent of school choice. In S. Sugarman & F. Kremerer (Eds.), School choice and social contro versy: Politics, policy and law (pp. 13–35). Washington, DC: Brookings Institution Press. Hess, F. (2002). Revolution at the margins: The impact of competition on urban school systems Washington, D.C.: Brookings. Hess, F., & Leal, D. (2001) Quality, race, and the urban education marketplace. Urban Affairs Review, 37 (2), 249–266. Hess, F., Maranto, R., & Milliman, S. (1999). Coping with competition: Ho w school systems respond to school choice (Occasional Paper 99–01). Camb ridge, MA: Harvard Program on Education Policy and Governance. Heubert, J. P., & Hauser R. M. (Eds.). (1999). High stakes: Testing for tracking, promotion, and graduation Washington, DC: National Academy Press. Howe, K., Eisenhart, M., & Bete benner, D. (2001). School choi ce crucible: A case study of Boulder Valley. Phi Delta Kappan 83 (2), 137–146. Hoxby, C. (2001). How school choice affects the achi evement of public school students (Tech. Rep.). Stanford, CA: Hoover Institution. (Paper pr epared for Koret Task Force Meeting on K– 12 Education) Lauder, H., & Hughs, D. (1999). Trading in futures: Why mark ets in education don’t work Philadelphia: Open University Press. Linn, R. L. (2000). Assess ments and accountability. Educational Researcher, 29 4–16. Linn, R. L. (2001). Reporting school quality in stan dards-based accountability systems (CRESST Policy Brief No. 3). Los Angeles, CA: Nation al Center for Research on Evaluation, Standards and Student Testing. Merrifield, J. (2001). The school choice wars Lanhan, MD: Scarecrow Press Inc. National Commission on Excelle nce in Education. (1983). A nation at risk: The imperative for educational reform (Tech. Rep.). Washington, DC: U.S. Government Printing Office. National Research Council. ( 1999). Accountab ility. In R. El more & R. Rothman (Eds.), Testing, teaching, and learning: A guide for states and school districts (pp. 91–101). Washington DC: National Academy Press. No Child Left Behind Act of 200 1. (2002). Public Law 107–110. Office of Technology Assessment. (1992). Testing in American schools: Asking the right questions (Tech. Rep. No. OTA–SET–519). Washin gton DC: U.S. Government Printing Office.


Education Policy Analysis Archives Vol. 13 No. 41 18 Orlofsky, G. F., & Olson, L. (2001). The state of the states. In Quality county 2001: A better balance (Vol. 20, pp. 86–88). Education Week. Reardon, S., & Yun, J. (2002). Private school racial enrollments and segregation In Kahlenberg, R.D. (Ed.) Public School Choice vs. Private School Vouchers New York: Century Foundation Press. Rothstein, L. (1999). School choice and students with disabilities. In S. Sugarman & F. Kremerer (Eds.), School choice and social controve rsy: Politics, policy and law (pp. 332–364). Washington, DC: Brookings Institution Press. Singer, J. D., & Will ett, J. B. (2003). Applied longitudinal data analysis New York: Oxford University Press. Snijders, T., & Bosker, R. (1999). Multilevel analysis Thousand Oaks, CA: Sage Publications. Wells, A. S. (1993). Time to choose: America at the crossroads of school choice policies. New York: Hill and Wang. Wells, A. S. (1998). Beyond the rhetoric of charter school re form: A study of ten California districts (Tech. Rep.). Los Angeles: Univer sity of California, Los Angeles. Wells, A. S., & Crane, R. L. (1 992). Do parents choose school qu ality or status? a sociological theory of free market educati on. In P. W. Cookson (Ed.), The choice controversy (pp. 65– 82). Newbury Park, CA : The Corwin Press. Witte, J. (2001). The market approach to education: An anal ysis of America’s first voucher program Princeton, NJ: Prince ton University Press. Zollers, N. (2000). Schools need rules when it comes to students with disabilities. Education Week, 19 (5), 46.


On School Choice and Test-Based Accountability 19 About the Authors Damian W. Betebenner Boston College Kenneth R. Howe University of Colorado at Boulder Samara S. Foster University of Colorado at Boulder Email: Damian W. Betebenner is assistant professor in th e Department of Educational Research, Measurement, and Evaluation. His research interests includ e longitudinal data analysis and applied statistics—particularly as they pertain to education policy. Kenneth R. Howe is professor in the Educational Foundations, Policy, and Practice program area. He specializes in education policy, profession al ethics, and philosophy of education. Samara S. Foster is a doctoral student in the School of Education at the University of Colorado at Boulder. She is in the Research and Evaluation Method ology and Educational Foundations, Policy, and Practice programs. Her research interests include philosophy of education, philosophy of social science rese arch, education policy, and feminist theory.


Education Policy Analysis Archives Vol. 13 No. 41 20 EDUCATION POLICY ANALYSIS ARCHIVES Editor: Sherman Dorn, University of South Florida Production Assistant: Chris Murre ll, Arizona State University General questions about ap propriateness of topics or particular articles may be addressed to the Editor, Sherman Dorn, Editorial Board Michael W. Apple University of Wisconsin David C. Berliner Arizona State University Greg Camilli Rutgers University Casey Cobb University of Connecticut Linda Darling-Hammond Stanford University Mark E. Fetler California Commission on Teacher Credentialing Gustavo E. Fischman Arizona State Univeristy Richard Garlikov Birmingham, Alabama Gene V Glass Arizona State University Thomas F. Green Syracuse University Aimee Howley Ohio University Craig B. Howley Appalachia Educational Laboratory William Hunter University of Ontario Institute of Technology Patricia Fey Jarvis Seattle, Washington Daniel Kalls Ume University Benjamin Levin University of Manitoba Thomas Mauhs-Pugh Green Mountain College Les McLean University of Toronto Heinrich Mintrop University of California, Berkeley Michele Moses Arizona State University Anthony G. Rud Jr. Purdue University Michael Scriven Western Michigan University Terrence G. Wiley Arizona State University John Willinsky University of British Columbia


On School Choice and Test-Based Accountability 21 EDUCATION POLICY ANALYSIS ARCHIVES English-language Graduate -Student Editorial Board Noga Admon New York University Jessica Allen University of Colorado Cheryl Aman University of British Columbia Anne Black University of Connecticut Marisa Burian-Fitzgerald Michigan State University Chad d'Entremont Teachers College Columbia University Carol Da Silva Harvard University Tara Donahue Michigan State University Camille Farrington University of Illinois Chicago Chris Frey Indiana University Amy Garrett Dikkers University of Minnesota Misty Ginicola Yale University Jake Gross Indiana University Hee Kyung Hong Loyola University Chicago Jennifer Lloyd University of British Columbia Heather Lord Yale University Shereeza Mohammed Florida Atlantic University Ben Superfine University of Michigan John Weathers University of Pennsylvania Kyo Yamashiro University of California Los Angeles


Education Policy Analysis Archives Vol. 13 No. 41 22 Archivos Analticos de Polticas Educativas Associate Editors Gustavo E. Fischman & Pablo Gentili Arizona State University & Universidade do Estado do Rio de Janeiro Founding Associate Editor for Spanish Language (1998—2003) Roberto Rodrguez Gmez Editorial Board Hugo Aboites Universidad Autnoma Metropolitana-Xochimilco Adrin Acosta Universidad de Guadalajara Mxico Claudio Almonacid Avila Universidad Metropolitana de Ciencias de la Educacin, Chile Dalila Andrade de Oliveira Universidade Federal de Minas Gerais, Belo Horizonte, Brasil Alejandra Birgin Ministerio de Educacin, Argentina Teresa Bracho Centro de Investigacin y Docencia Econmica-CIDE Alejandro Canales Universidad Nacional Autnoma de Mxico Ursula Casanova Arizona State University, Tempe, Arizona Sigfredo Chiroque Instituto de Pedagoga Popular, Per Erwin Epstein Loyola University, Chicago, Illinois Mariano Fernndez Enguita Universidad de Salamanca. Espaa Gaudncio Frigotto Universidade Estadual do Rio de Janeiro, Brasil Rollin Kent Universidad Autnoma de Puebla. Puebla, Mxico Walter Kohan Universidade Estadual do Rio de Janeiro, Brasil Roberto Leher Universidade Estadual do Rio de Janeiro, Brasil Daniel C. Levy University at Albany, SUNY, Albany, New York Nilma Limo Gomes Universidade Federal de Minas Gerais, Belo Horizonte Pia Lindquist Wong California State University, Sacramento, California Mara Loreto Egaa Programa Interdisciplinario de Investigacin en Educacin Mariano Narodowski Universidad To rcuato Di Tella, Argentina Iolanda de Oliveira Universidade Federal Fluminense, Brasil Grover Pango Foro Latinoamericano de Polticas Educativas, Per Vanilda Paiva Universidade Estadual Do Rio De Janeiro, Brasil Miguel Pereira Catedratico Un iversidad de Granada, Espaa Angel Ignacio Prez Gmez Universidad de Mlaga Mnica Pini Universidad Nacional de San Martin, Argentina Romualdo Portella do Oliveira Universidade de So Paulo Diana Rhoten Social Science Research Council, New York, New York Jos Gimeno Sacristn Universidad de Valencia, Espaa Daniel Schugurensky Ontario Institute for Studies in Education, Canada Susan Street Centro de Investigaciones y Estudios Superiores en Antropologia Social Occidente, Guadalajara, Mxico Nelly P. Stromquist University of Southern California, Los Angeles, California Daniel Suarez Laboratorio de Politicas Publicas-Universidad de Buenos Aires, Argentina Antonio Teodoro Universidade Lusfona Lisboa, Carlos A. Torres UCLA Jurjo Torres Santom Universidad de la Corua, Espaa


Download Options

Choose Size
Choose file type
Cite this item close


Cras ut cursus ante, a fringilla nunc. Mauris lorem nunc, cursus sit amet enim ac, vehicula vestibulum mi. Mauris viverra nisl vel enim faucibus porta. Praesent sit amet ornare diam, non finibus nulla.


Cras efficitur magna et sapien varius, luctus ullamcorper dolor convallis. Orci varius natoque penatibus et magnis dis parturient montes, nascetur ridiculus mus. Fusce sit amet justo ut erat laoreet congue sed a ante.


Phasellus ornare in augue eu imperdiet. Donec malesuada sapien ante, at vehicula orci tempor molestie. Proin vitae urna elit. Pellentesque vitae nisi et diam euismod malesuada aliquet non erat.


Nunc fringilla dolor ut dictum placerat. Proin ac neque rutrum, consectetur ligula id, laoreet ligula. Nulla lorem massa, consectetur vitae consequat in, lobortis at dolor. Nunc sed leo odio.