high risk, low return: the achievement effects of restructuring in disadvantaged schools

15
Social Psychology of Education 3: 245–259, 2000. c 2000 Kluwer Academic Publishers. Printed in the Netherlands. 245 High risk, low return: The achievement effects of restructuring in disadvantaged schools ROGER C. SHOUSE ? and LAWRENCE J. MUSSOLINE The Pennsylvania State University Pine Grove Area School District (Received: 19 October 1999; in final form: 29 November 1999) Abstract. Studies of school ‘restructuring’ have linked the concept to a specific normative agenda of organizational and instructional practices (such as individualized or small-group instruction, het- erogeneous grouping, and team teaching). Though a number of theoretical and empirical studies point to the positive direct and indirect effects of these practices on student achievement, they tend to overlook a body of prior work suggesting that students in disadvantaged schools tend to benefit academically from more traditional types of arrangements. Tracing and building upon this literature, this study examines the impact of ‘restructuring’ on mathematics achievement among students attending a sample of socioeconomically disadvantaged schools. After controlling for salient student background characteristics, our study found no significant difference in math achievement levels between nonrestructured schools and schools that had been restructured for at least three years. However, significantly lower levels of math achievement were found among schools that had been ‘restructured’ for less than three years. We argue that these findings reflect the fact that the high complexity associated with many ‘restructured’ practices pose special problems and risks for disadvantaged schools. 1. Introduction The idea of the ‘restructured school’ has attracted considerable attention from educators and school policymakers over the past decade. Represented by such organizational and instructional practices as heterogeneous grouping, cooperative learning, and cross-disciplinary team teaching, the idea has gathered considerable normative momentum in American secondary schools. Put another way, as school restructuring has become associated with a specific agenda of reforms, it has at- tained ‘movement’ status. The movement’s main conceptual thrust seems to be that traditional forms of schooling, overly bureaucratic in nature, tend to alienate students from meaningful learning experience and that this problem can be alle- viated by shifting to a more collegial, cooperative, and communal organizational model (Center on Organization and Restructuring of Schools, 1992; Newmann & Wehlage, 1995). ? Address for correspondence: Roger Shouse, Penn State, College of Education, 302B Rackley Building, University Park, PA 16802; Tel.: 814-863-3773; Fax: 814-865-1480; E-mail: [email protected]

Upload: roger-c-shouse

Post on 03-Aug-2016

213 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: High risk, low return: The achievement effects of restructuring in disadvantaged schools

Social Psychology of Education3: 245–259, 2000.c© 2000Kluwer Academic Publishers. Printed in the Netherlands. 245

High risk, low return: The achievement effects ofrestructuring in disadvantaged schools

ROGER C. SHOUSE? and LAWRENCE J. MUSSOLINEThe Pennsylvania State UniversityPine Grove Area School District

(Received: 19 October 1999; in final form: 29 November 1999)

Abstract. Studies of school ‘restructuring’ have linked the concept to a specific normative agendaof organizational and instructional practices (such as individualized or small-group instruction, het-erogeneous grouping, and team teaching). Though a number of theoretical and empirical studiespoint to the positive direct and indirect effects of these practices on student achievement, theytend to overlook a body of prior work suggesting that students in disadvantaged schools tend tobenefit academically from more traditional types of arrangements. Tracing and building upon thisliterature, this study examines the impact of ‘restructuring’ on mathematics achievement amongstudents attending a sample of socioeconomically disadvantaged schools. After controlling for salientstudent background characteristics, our study found no significant difference in math achievementlevels between nonrestructured schools and schools that had been restructured for at least threeyears. However, significantly lower levels of math achievement were found among schools that hadbeen ‘restructured’ for less than three years. We argue that these findings reflect the fact that thehigh complexity associated with many ‘restructured’ practices pose special problems and risks fordisadvantaged schools.

1. Introduction

The idea of the ‘restructured school’ has attracted considerable attention fromeducators and school policymakers over the past decade. Represented by suchorganizational and instructional practices as heterogeneous grouping, cooperativelearning, and cross-disciplinary team teaching, the idea has gathered considerablenormative momentum in American secondary schools. Put another way, as schoolrestructuring has become associated with a specific agenda of reforms, it has at-tained ‘movement’ status. The movement’s main conceptual thrust seems to bethat traditional forms of schooling, overly bureaucratic in nature, tend to alienatestudents from meaningful learning experience and that this problem can be alle-viated by shifting to a more collegial, cooperative, and communal organizationalmodel (Center on Organization and Restructuring of Schools, 1992; Newmann &Wehlage, 1995).

? Address for correspondence: Roger Shouse, Penn State, College of Education, 302B RackleyBuilding, University Park, PA 16802; Tel.: 814-863-3773; Fax: 814-865-1480; E-mail: [email protected]

Page 2: High risk, low return: The achievement effects of restructuring in disadvantaged schools

246 ROGER C. SHOUSE AND LAWRENCE J. MUSSOLINE

This argument has been explored in various ways, and one prominent studyhas reported significant achievement effects associated with being a ‘restructuredschool’ (Lee & Smith, 1995). One major concern left largely unexamined in re-structuring literature, however, is whether students in all types of schools standto benefit from the restructuring reform agenda. For example, schools often differdramatically in the availability of human and social resources necessary to success-fully implement reform practices. Students may also differ substantially in theirneed for and response to different instructional approaches. Specifically, a longstanding body of research links positive achievement effects to more traditionalorganizational characteristics and instructional practices in socioeconomically dis-advantaged schools (see, Hallinger & Murphy, 1986; Pogrow, 1997; Shouse, 1998,for example).

Drawing on data from the National Education Longitudinal Study (NELS:88)tenth-grade school and student surveys, this paper examines the impact of schoolrestructuring on the mathematics achievement scores of students attending so-cioeconomically disadvantaged schools. In contrast to the claims of some restruc-turing advocates, its findings indicate that substantial negative achievement effectsmay occur when these schools adopt of a set of restructured practices. Whileseveral explanations for this finding are discussed, we aim special focus at thepossibility that the complexity of many reform practices poses greater demandsand risks for disadvantaged schools than for their more affluent counterparts.

2. Theoretical framework

‘Restructuring’ is basically conceived as representing a purposeful change in pat-terns of rules, roles, and relationships within schools (Corbett, 1990). More pre-cisely, this change is said to unfold along three dimensions. First, there is structuralchange as schools shift away from centralized, bureaucratic, and rigid forms ofgovernance (Darling-Hammond, 1995; Murphy, 1991). Second, there is profes-sional change as teachers begin to act more collegially; assume more responsibilityfor student learning, and assume more personalistic and caring attitudes towardstudents (Louis, Marks & Kruse, 1996; Murphy, 1991). Finally, there is instruc-tional change as focus shifts from individual to cooperative activities and fromaccumulating specific knowledge to developing ‘thinking skills’ (Elmore, 1990;Murphy, 1991).

Alternatively, one may think about restructuring in terms of its contrast withthe so-called ‘effective schools’ model. The latter stresses the need to intensify orexpand the influence of traditional school practices and characteristics (more strin-gent course requirements, common academic curriculum, ‘direct instruction’, etc.).Restructuring, on the other hand, calls upon schools to create more communal, par-ticipatory arrangements. Thus, although it never uses the term ‘restructuring’, Brykand Driscoll’s (1988) study of the positive effects of school communal organization(communality) can thus be viewed as an early example of restructuring research.

Page 3: High risk, low return: The achievement effects of restructuring in disadvantaged schools

RESTRUCTURING IN DISADVANTAGED SCHOOLS 247

Though the student outcomes associated with restructuring tend to vary in theliterature (Elmore, 1990), at the heart of the movement lies a belief that it willhelp raise student achievement (Newmann, 1992). Testing this hypothesis, how-ever, calls on researchers to operationalize the concept of the restructured school insome satisfactory way so that its achievement effects can be empirically measured.In Bryk and Driscoll’s study cited above, for example, items from the High Schooland Beyond study were molded into an index of school communal organization.Based on hierarchical regression modeling, Bryk and Driscoll reported a positiverelationship between communality and student mathematics achievement.

Building on the conceptual link between communality and student achievement,Lee and Smith’s (1995) subsequent study categorized schools as restructured basedon their having implemented at least three (of 12) practices thought to signify aschool’s shift away from bureaucratic organization (as indicated primarily throughprincipal reports; see Table I for a list of these practices). Restructured schoolswere not only found to have higher levels of student achievement but also smallergaps in achievement between students of different socioeconomic backgrounds.

Yet, of the 12 restructured practices listed in Table I, only four directly relate toclassroom instruction (i.e., instructional teams, independent study, heterogeneousgrouping, and cooperative learning). And though several of the items listed inTable I bear plausible connection to student achievement, empirical evidence asto their impact tends to be mixed at best. It is important to note, however, thatLee and Smith conceive an indirect link between these kinds of practices andstudent achievement. Their adoption is said to reflect a school’s effort to createa less bureaucratic, more communal and collegial, culture and climate.

But while using lists or sets of practices to operationalize the concept of re-structuring may be useful to researchers, it can be problematic for educational

Table I. NELS:88 first follow-up survey items representingrestructured practices

Students keep the same homeroom through high school

Emphasis on staff solving school problems

Parents volunteer in school

Use of Interdisciplinary teaching teams

Independent study in math and science

Independent study in English and social studies

Mixed-ability classes in math and science

Cooperative learning focus

Students’ evaluation of teachers important

School within a school

Teacher teams have common planning time

Flexible time for classes

From Lee and Smith, 1995.

Page 4: High risk, low return: The achievement effects of restructuring in disadvantaged schools

248 ROGER C. SHOUSE AND LAWRENCE J. MUSSOLINE

policy and school improvement. Consider, for example, the practice of team teach-ing across content areas. Though based on reasoned speculation concerning theneed for knowledge integration, little if any evidence exists linking the practiceto higher student achievement. In addition, cross content team teaching is an ex-tremely complex form of instructional technology, requiring considerable commit-ment and coordination of teachers’ time and energy. Math and science teachers, inparticular, often raise concerns about the instructional gaps that result from havingto sacrifice content coverage in favor of thematic congruence. (See, Meister, 1997,for an excellent case study of this practice.) When ‘team teaching’ gets packagedinto the restructuring agenda, however, local school or district policymakers cometo view the practice as an important way to demonstrate their interest in change andimprovement. A sense of normative momentum – a ‘bandwagon effect’ – developsthat can overshadow practical considerations or sound empirical evidence.

Indeed, there is substantial cause for educators and policymakers to harborhealthy skepticism about some school reform proposals. Organizational charac-teristics or instructional practices linked to effectiveness in one school setting maybe ineffective, impractical, or even risky in others. For example, Hallinger andMurphy’s (1986) study comparing effective low- and high-SES schools found taskorientation and tight curricular and instructional control to be characteristic ofprincipals in effective low-SES schools. In effective high-SES schools, however,principals tended to be more ‘relationship oriented’ and allowed their teachersmore autonomy in exercising professional judgments. While effective high-SESprincipals sought strong ties with their surrounding communities, their counterpartsin low-SES schools acted as gatekeepers, insulating their school from externalinfluences that might interfere with academic goals. In short, leadership in effect-ive low-SES schools tended to be more bureaucratic and traditional; in effectivehigh-SES schools it appeared more communal and collegial.

Coupled with Hallinger and Murphy’s findings is notable evidence from the‘effective schools’ studies of the 1980s. The ‘effective practices’ identified withinthat literature not only tended to represent a more traditional and intensified ap-proach to teaching and learning (Elmore, 1990) but also tended to be based onevidence from urban elementary schools (Good & Brophy, 1986). A more recentstudy of the achievement effects of high school academic press and communalorganization also suggests that the achievement effects of restructuring might differacross categories of school urbanicity or socioeconomic status (Shouse, 1996a).Based on data from the National Education Longitudinal Study (NELS:88), theresults indicated that students in low-SES schools tended to benefit the most fromacademic press, a relatively traditional organizational characteristic. Strong com-munality was actually linked to significantly lower achievement for students inlow-SES schools with weak academic press. In high-SES schools, however, theimpact of communality was independent of, and actually outweighed, the impactof academic press.

Page 5: High risk, low return: The achievement effects of restructuring in disadvantaged schools

RESTRUCTURING IN DISADVANTAGED SCHOOLS 249

A subsequent study of middle and junior high schools (Shouse, 1998) also re-ported a number of contrasts between the effects of traditional and restructuredpractices across urban and nonurban schools. Specifically, lower levels of mathachievement were found in urban schools using cooperative learning and/or middleschool grade configuration. In addition, urban schools’ use of such traditional prac-tices as ability grouping in mathematics and the establishment of formal policieshighlighting the importance of daily homework were found tied to positive effects.Studies such as these call on us to rethink our perceptions about the comparativevalues of traditional and reform practices across school socioeconomic settings.

Aside from this empirical evidence, however, a conceptual basis also exists forsuspecting that such contextual differences affect students’ academic needs. Onefactor relates to the ability of more advantaged schools to tap into and take advant-age of their students’ cultural and social capital (Hirsch, 1996; Shouse, 1996a, b).First, for example, more affluent students are more likely to have accrued higherlevels of academic knowledge (either from within or outside of school), and thisraises the likelihood that some types of restructured classroom practices will bemore effective (e.g., those emphasizing student centered learning and the acquisi-tion of higher order skills). Second, the denser, more academically oriented socialsupport structures available to families in more advantaged communities help re-inforce and facilitate school academic goals. A ‘safety net’ thus exists for moreadvantaged schools that reduces the potential negative impact of flawed, poorlyexecuted, or highly complex instructional practices.

Lacking this safety net, the use of restructuring reforms constitutes a muchriskier proposition for disadvantaged schools. Conveying this idea of risk quiteclearly, Patricia Graham’s metaphor regarding progressive and traditional school-ing suggests the dilemma they face in selecting instructional technologies. Grahamasks us to consider

...progressive education as pork chops, and traditional education as apples

.... A pork chop at it’s best is absolutely wonderful and is enhanced by theaddition of a baked apple. But the pork chop half done, half cooked ... willgive you trichinosis. The baked apple half cooked is not very imaginative, butis still very nutritious. (1985, p. 12)

Thus, the positive effects of certain traditional practices and negative effects ofcertain restructured practices in disadvantaged schools may relate to the differencesin the ‘culinary skill’ of their teachers and the ‘nutritional needs’ of their students.(The notion of technological risk is discussed further in the final section of thispaper.)

An additional problem involves the potential for schools shifting toward morecommunal, caring, and personalistic structures and relationships to also shift theirfocus to emphasize social over academic goals. This concern becomes larger inrelation to the degree of social challenge students face. A high school teacherinterviewed by one of the present authors, for example, described how gangs had

Page 6: High risk, low return: The achievement effects of restructuring in disadvantaged schools

250 ROGER C. SHOUSE AND LAWRENCE J. MUSSOLINE

been threatening one of his students. He wondered how he could expect this studentto learn the academic course material when people outside the school were tryingto kill him. This is an extreme example, of course, but it does suggest how schools’goals and classroom activities can become more oriented toward safety, security,and good behavior than toward academic activity. Is it not reasonable to speculatethat the more bureaucratic structures and relationships typical of traditional schoolsmight help preserve the emotional distance necessary for effective teaching tooccur?

3. Methodology

Our study had a modest goal which related to the materials reviewed above, namely,to examine the extent to which the positive effects of restructuring on students’math achievement reported in Lee and Smith’s (1995) study applied also to asubsample of disadvantaged schools. Starting with their operational definition of arestructured school – those which had adopted at least three of the 12 restructuredpractices represented in the NELS:88 10th-grade school survey – we sought alsoto discover the extent to which these achievement effects might vary in relation tothe amount of time the reforms had actually been used. To this end, our analysisdistinguished between schools that had employed restructured practices for at leastthree years and those that had implemented them more recently.

3.1. SOURCE OF THE DATA

Our analyses are based on data from the NELS:88 First Follow-Up (10th-grade)survey. Conducted by the National Center for Education Statistics, NELS:88 ex-amined students’ educational experiences from 8th grade through high school andinto college, highlighting family, community, school, and classroom factors thatinfluenced educational success.

Beginning with a Base Year (8th-grade) survey, the NELS:88 study also in-cluded follow-up surveys when students were in the 10th grade and 12th grade,respectively, as well as a third follow-up survey which surveyed students two yearsafter their high school graduation. To obtain a representative sample of Americanpublic and private school students, the Base Year survey began with a represent-ative sample of 1,035 8th-grade public and private schools stratified by region,urbanicity, and minority enrollment. A random sample of students was obtained ineach school, resulting in an overall student sample size of approximately 24,599students. In addition to a questionnaire, each student in the Base Year survey com-pleted a cognitive test in history, mathematics, reading, and science. Questionnaireswere also administered to a parent, the principal, and two teachers for each student(one from science or math, the other from English or social studies).

The NELS:88 First Follow-Up survey (conducted in 1990) had the same basiccomponents and student sample. However, as students in the Base Year sample

Page 7: High risk, low return: The achievement effects of restructuring in disadvantaged schools

RESTRUCTURING IN DISADVANTAGED SCHOOLS 251

were followed to their tenth-grade schools, some respondents fell out of the sample,and researchers examining school effects had to grapple with three methodologicalissues. First, though the student sample was still nationally representative, therewas no longer an assurance of a random student sample within each high school.Second, there was no longer assurance that the school sample was nationally rep-resentative. Finally, within many schools, the student and teacher sample sizes nolonger permitted a comprehensive analysis of school effects.

To address these issues, our analyses used a subsample of socioeconomicallydisadvantaged first follow-up schools having at least five NELS-sampled students.Fairly common in studies using NELS:88 data, this subsampling technique in-creased the likelihood of obtaining a representative student sample within eachschool and facilitated the use of the hierarchical regression analysis necessary foraccurate estimates of school effects. Schools were identified as socioeconomicallydisadvantaged if their school mean SES was at least one standard deviation belowthat for the entire NELS:88 10th-grade school sample. The resulting subsampleconsisted of 2,113 students across 140 public and private high schools. (This com-pares with an overall sample of 14,943 students across 742 schools fitting thebase-level, ‘at-least-five-students-per-school’ criterion.)

3.2. GENERAL ANALYSIS STRATEGY

The achievement effect of restructuring among this sample of low-SES schoolswas examined using hierarchical linear modeling (HLM), a type of multiple re-gression technique useful in examining student and school factors associated withvariation in achievement. Using HLM, the total variation in student achievementwas separated into its student- and school-level components. Separate regressionequations (or ‘models’) were then specified to explain the variation at each level.The intercept of the student-level equation represented the estimated grand meanachievement across the sampled schools, adjusted for the specified student-levelvariables. The individual school means comprising this grand mean served as thedependent variable of the school-level equation (which links school-level variablesto variation in student achievement across schools).

In our analysis, which focused primarily on the achievement effects of school-level characteristics, the student-level model served mainly as a set of controls;that is, to reduce the likelihood that any observed school-level effects might actu-ally result from variation in student composition across schools. For example, ourstudent-level model included variables representing student SES, race, high schoolprogram, prior course work in math, and prior academic ability. Table II presentsdescriptive statistics for all variables used in this analysis. (For comparative pur-poses, descriptive statistics are also presented for the nondisadvantaged NELS:88schools not included in this analysis.)

Our school-level models included four variables. One was a continuous measureof school-average SES based on the average SES of the NELS-sampled students

Page 8: High risk, low return: The achievement effects of restructuring in disadvantaged schools

252 ROGER C. SHOUSE AND LAWRENCE J. MUSSOLINE

Table II. Variable descriptions and simple statistics

Description NELS source Disadvantage Other schools∗variable schools Mean SD

Mean SD

Student level

10th-grade standardized math IRT score F12XMIRR 37.82 12.44 46.60 13.77

8th-grade standardized math IRT score BY2XMIRR 31.43 9.92 38.84 12.10

Semesters of geometry since 9th grade F1S22SUM 0.69 0.97 1.19 1.00

NELS:88 student SES composite F1SES −0.58 0.66 0.21 0.76

Asian student F1RACE 0.03 0.17 0.06 0.24

Hispanic student F1RACE 0.22 0.41 0.07 0.26

African–American student F1RACE 016 0.37 0.07 0.26

Student in academic program F1HSPRGM 0.28 0.45 0.39 0.49

Student in vocational program F1HSPRGM 0.32 0.47 0.19 0.39

School level

School mean SES MEANSES −1.29 0.34 0.00 1.00

Restructured school (see Table I) 0.32 0.47 0.34 0.47

Restructured for at least three years (see Table I) 0.16 0.37 0.17 0.37

Restructured within past three years (see Table I) 0.16 0.37 0.17 0.38

∗For comparative purposes only, not included in this analysis.

in each school. This helped to control for the relative impact of school affluencewithin our sample of disadvantaged schools. Three dichotomous, or ‘dummy’ vari-ables were also included. The first of these indicated whether or not a schoolqualified as restructured based on Lee and Smith’s (1995) criteria described earlierin this paper. That is, schools were considered restructured if their principals repor-ted the use of at least three restructured practices (see Table I) and nonrestructuredif they did not.1 The other two dummy variables distinguished between schools thathad been restructured for at least three years and those that had become restructuredwithin the three years prior to the survey. Structuring the variables this way enabledus to gauge whether restructuring effects on math achievement were influenced bythe duration of their use.

3.3. DEPENDENT VARIABLE

Student achievement was represented by the NELS:88 mathematics IRT (item re-sponse theory) scores (see National Center for Education Statistics, 1994). ThoughNELS:88 also included measuring for reading, history, and science achievement,the math scores represented the most valid indicator of school effectiveness. Forone thing, because the math test contained the greatest number of items and ability

Page 9: High risk, low return: The achievement effects of restructuring in disadvantaged schools

RESTRUCTURING IN DISADVANTAGED SCHOOLS 253

Table III. Effects of school restructuring on math achievement indisadvantaged schoolsa

Variable Coeff. Coeff.

School level

School mean math achievement (intercept)−0.50∗ −0.50∗School mean SES 0.09∗ 0.09∗All restructured schools −0.14∗ –

Schools restructured for at least three years −0.04

Schools restructured within past three years −0.25∗

Student level

8th-grade math achievement 0.82∗ 0.82∗Semesters of geometry 0.08∗ 0.08∗Student SES 0.03∗ 0.03∗Asian student −0.15∗ −0.15∗Hispanic student −0.04 −0.04

Black student −0.12∗ −0.12∗Student in academic program 0.13∗ 0.13∗Student in vocational program −0.13∗ −0.13∗

∗Coefficient statistically significant atp < 0.05.aNinety percent of the total variance in student math achievement iswithin schools, 10% is between schools. The student-level variablesincluded here account for 76% of the within-school variance in mathachievement. The school level variables account for a total of 22% ofthe total between-school variance, and school-mean SES accounts for72% of this explained variance.

levels, it was most immune to floor and ceiling effects. In addition, math scoreswere more likely to reflect in-school learning than other subject scores (Haney,1996; Karweit & Ricciuti, 1997). (For additional technical information regardingNELS:88 test validity and reliability, see Rock, Pollack & Quinn, 1995).

4. Results

Table III displays the results of our analysis. As is customary in presenting HLMresults, school-level effects are presented in the top table panel, student-level ef-fects are represented in the bottom. The coefficients are presented as ‘effect sizes’;that is, they represent the expected difference (measured in standard deviations) instudent math achievement associated with a ‘one-unit’ change in the independentvariable. For some variables (mean SES, 8th-grade math achievement, and studentSES), one-unit equals one standard deviation. For dichotomous variables (thoserepresenting school restructuring status, student race, and student program) thecoefficient represents simply the expected difference between members and non-

Page 10: High risk, low return: The achievement effects of restructuring in disadvantaged schools

254 ROGER C. SHOUSE AND LAWRENCE J. MUSSOLINE

members of the particular category. The student-level variables served primarily ascontrols; that is, their inclusion decreased the chance that any school-level effectsreported here actually resulted from cross-school variation in student backgroundand ability. In reporting results, the student variables have been centered upon amean of zero. This centering does not affect how coefficients are interpreted, andit allows the school effects and intercept reported in the top panel to reflect theexperience of an average general track student.2

Two school-level models are presented across the two columns of Table III. Thefirst model indicates a significant negative effect(−0.14) associated with beinga restructured school, even after controlling for the effects of school affluenceand student background. In other words, at first glance, we would expect to findthat among disadvantaged schools, students in nonrestructured schools outperformtheir peers in restructured schools by 14% of a standard deviation. The secondcolumn, however, indicates this first reading to be somewhat misleading. After dis-tinguishing between longer-term and more recently restructured schools, Table IIIreveals (a) no significant difference between nonrestructured and longer-term re-structured schools and (b) a significant negative effect(−0.25) associated withbeing a recently restructured school (as compared with nonrestructured schools).

As math achievement scores are for all practical purposes normally distributedacross the entire NELS:88 student sample, we can translate this negative effect intopercentiles. Based on the intercept presented in Table III, and using a table ofZ

scores, we would expect a difference of about eight percentiles between a typicalgeneral program 10th-grade student in a recently restructured school and his or hercounterpart in a nonrestructured or longer-term restructured school.

5. Discussion

The results reported here reinforce our concern that local school policymakersneed to maintain some healthy skepticism about the usefulness of some of thesystemic now reforms being advocated in school restructuring literature. Not onlydid the disadvantaged schools in our sample accrue no long-term achievement ad-vantage, but those having recently adopted restructuring reforms appeared to lagsubstantially behind.

One additional note seems worthy of mention here. As part of our exploratoryanalyses, we examined the separate effects of the 12 practices classified as restruc-tured. These analyses revealed no distinctive negative effect or pattern of negativeeffect. Though it seems tempting to speculate that what we actually found wassome sort of ‘fireman effect’ (that is, just as firemen surround burning buildings,‘reforms’ tend to be found in poorly performing schools), we strongly suspect thatthis is not what is happening. For one thing, our exploratory analyses revealedno comparable negative effect associated with schools adopting three or moretraditional types of practices.

Page 11: High risk, low return: The achievement effects of restructuring in disadvantaged schools

RESTRUCTURING IN DISADVANTAGED SCHOOLS 255

Earlier in this paper, we offered two possible reasons as to why restructuringmight be problematic for student achievement in disadvantaged schools. One re-lated to the complexity, high resource demand, and risk of some restructuringreforms. The other involved the hazards which increased communality may pose toa sense of academic mission in these schools. Having discussed this latter problemin prior articles (Shouse, 1996a, b), we focus our attention here on the problems ofcomplexity, resource demand, and risk.

As previously discussed, it is evident that the agenda of the restructuring move-ment contains a number of instructional technologies that are relatively complex ortime consuming, often stretching the professional capacities of teaching staffs. Wehave already discussed the example of team teaching across content areas. The useof heterogeneous grouping and cooperative learning serves as two additional illus-trations of this problem. For example, heterogeneous grouping calls upon teachersto deliver appropriate instruction to a much wider range of student needs and abil-ities than is typically the case in homogeneous classes. And, although cooperativelearning is often prescribed as a means of mitigating this increased demand, to bedone well it also requires highly skilled teachers capable of maintaining effect-ive control over instructional pacing in and across the cooperative work groups.The type of classroom orchestration needed for successful cooperative learningrequires substantial training, something that often lies beyond the professionalor fiscal resources of low-SES schools. Other organizational characteristics withwhich schools serving disadvantaged communities must often struggle can makecooperative learning even more problematic. For example, students may be morelikely to lack the prerequisite knowledge, interest, or motivation necessary for in-dependent academic work; ability distributions within classrooms may constrainteachers’ from forming balanced ability groupings; teachers may view the practiceas more of a social than academic exercise, etc.

These remarks are not intended as a blanket criticism of instructional experi-mentation or reform. The point, however, is that as instructional practices becomemore complex, the potential risk to student learning increases, especially whenschools or students lack the resources critical to their successful implementation.To illustrate, we might think of student achievement as a function of (a) the avail-ability of academically oriented social capital outside of school and (b) teachers’instructional skill. Compared to students in disadvantaged communities, those inmore affluent communities tend to have greater access to academically orientedsocial capital, and their teachers tend to be more professionally skilled. Two resultsfollow. Not only will teachers in disadvantaged schools tend to be less preparedto effectively implement restructured practices but it is also less likely that theirstudents will have access to any strong ‘academic safety net’ outside the school.In other words, complexity and organizational constraints increase the potentialfor professional error in disadvantaged schools, and the student learning in theseschools will be far more sensitive to professional error.

Page 12: High risk, low return: The achievement effects of restructuring in disadvantaged schools

256 ROGER C. SHOUSE AND LAWRENCE J. MUSSOLINE

This combination of circumstances leads us to suggest that urban, low-SES,or otherwise disadvantaged schools may be thought of as ‘high-risk’ organiza-tions in that the task they face is complex and the consequences of failure canbe extremely dire for both students and society. Though some researchers andpolicymakers consider restructuring practices as a way of addressing task com-plexity and reducing risk, in fact the complexity of these technologies themselvesmay actually increase the risk. The substantial achievement gap we report herebetween newly restructured and other schools thus exemplifies something quitesimilar to the ‘normal accidents’ that can result when tightly coupled organiza-tions increase their technological complexity (Perrow, 1999). That schools tendto be more loosely coupled might explain the absence of any gap between non-restructured and longer-term restructured schools. Over time, loose organizationalconnections allow teachers to modify or even circumvent reform practices to suittheir needs, resources, and skills. Alternatively, the gradual disappearance of anyachievement difference between traditional and restructured schools may result asteachers improve their skills over time or as students adjust to the new teachingtechniques.

Having pointed out the difficulties particular types of restructuring reforms posefor disadvantaged schools, it is important to emphasize that our analysis takesissue – not with restructuring per se – but with interpreting or measuring theconcept in terms of a single list of reforms. Ironically, this sort of ‘agendizing’can lead to administrative imposition of restructuring on professional staffs. Forexample, an urban high school principal faced with teacher skepticism regardinga district-proposed ‘empowerment’ plan is reported to have stated angrily to herstaff, “People, understand this! We will become an empowered school!” (Boyd &Shouse, 1997, p. 154).

In truth, school restructuring ought not be viewed as just one thing or as oc-curring in just one direction. In response to the structural constraints they face,people who work in schools, be they administrators or teachers, formal or informalleaders, hold varying and valuable views regarding the most effective paths toimprovement. Thus, for some, restructuring might mean striving for greater pro-fessional collegiality or decentralized decisionmaking. For others, it might meanadopting or incrementally improving upon traditional instructional practices. Forexample, instead of total detracking, a school may set out to reduce the numberof ability groups, infuse lower-ability classes with more ‘high-status’ curricularcontent, or assign their best teachers to such classes. Instead of rushing to imple-ment formal team teaching across content areas, a school may create conditionsthat foster teacher professional cooperation especially within content areas. Someschool leaders, rather than focus on specific instructional practices, may opt to be-gin a schoolwide discussion about core values or finding ways to increase academicpress.

Our analysis seeks not to unduly denigrate the practices included in the restruc-turing agenda or to cast undue praise on those they might replace. Instead, our point

Page 13: High risk, low return: The achievement effects of restructuring in disadvantaged schools

RESTRUCTURING IN DISADVANTAGED SCHOOLS 257

is that if restructuring is to be useful, researchers and educators ought not view itas the ‘one true path’ to the good and caring school. We thus agree heartily withLarry Cuban (1998) when he notes that school ‘goodness’ can be found in both tra-ditional and progressive settings. Our evidence supports his view and suggests theneed for policymakers, especially those concerned about improving disadvantagedschools, to think more critically and incrementally about school restructuring andreform.

Acknowledgements

This research was supported by a grant from the American Educational ResearchAssociation which receives funds for its ‘AERA Grants Program’ from the Na-tional Science Foundation, the National Center for Education Statistics, and theOffice of Educational Research and Improvement (U.S. Department of Education)under NSF Grant #RED-9452861. Opinions reflect those of the authors and do notnecessarily reflect those of the granting agencies.

Notes

1. In addition to identifying ‘restructured’ schools, Lee and Smith (1995) also identified what theyreferred to as ‘moderate’ and ‘unrestructured’ schools, based on their use of other types of prac-tices and reforms.

2. See Bryk and Raudenbush (1992, pp. 25–28) for further details.

References

Boyd, William L. & Shouse, Roger C. (1997). The problems and promise of urban schools. In H. J.Walberg, O. Reyes, & R. Weissberg (Eds),Children and youth: Interdisciplinary perspectives.Thousand Oaks, CA: Sage Publications, pp. 141–165.

Bryk, Anthony S. & Driscoll, Mary E. (1988).The school as community: Theoretical foundations,contextual influences, and consequences for students and teachers. Chicago: The University ofChicago Benton Center for Curriculum and Instruction.

Bryk, Anthony S. & Raudenbush, Stephen W. (1992).Hierarchical linear models: Applications anddata analysis methods. Newbury Park, CA: Sage Publications.

Center on Organization and Restructuring of Schools (1992).Estimating the extent of schoolrestructuring. Issues in restructuring schools, Fall, 1992. Madison, WI: Author.

Corbett, H. D. (1990).On the meaning of restructuring. Philadelphia: Research for Better Schools.Cuban, Larry (1998). A tale of two schools.Education Week, 28, 48.Darling-Hammond, Linda (1995). Restructuring schools for student success.Daedalus, 124(4), 153–

162.Elmore, Richard F. (1990). On changing the structure of public schools. In R. F. Elmore (Ed.),

Restructuring schools: The next generation of educational reform. San Francisco: Jossey-Bass,pp. 1–28.

Good, Thomas L. & Brophy, Jere E. (1986). School effects. In M. C. Wittrock (Ed.),Handbook ofresearch on teaching. New York: Macmillan.

Page 14: High risk, low return: The achievement effects of restructuring in disadvantaged schools

258 ROGER C. SHOUSE AND LAWRENCE J. MUSSOLINE

Graham, Patricia A. (1985). Schools: The raw, the cooked, and the half-baked.MassachusettsAssociation of School Committees Journal, 18(3), 12–17.

Hallinger, Phillip & Murphy, Joseph F. (1986). The social context of effective schools.AmericanJournal of Education, 94(3), 328–355.

Haney, Walt (1996). Reflecting on authentic assessment and school reform.Teachers College Record,98(2), 329–345.

Hirsch, E. D. (1996).The schools we need. New York: Doubleday.Karweit, Nancy & Ricciuti, Anne E. (1997).Summer learning: Achievement profiles of prospects

first graders. Washington, DC: U.S. Department of Education, Planning and Evaluation Service.Lee, Valerie E. & Smith, Julia B. (1995). Effects of high school restructuring and size on early gains

in achievement and engagement.Sociology of Education, 68(4), 241–270.Louis, Karen S., Marks, Helen M., & Kruse, Sharon (1996). Teachers’ professional community in

restructuring schools.American Educational Research Journal, 3(4), 757–798.Meister, Denise G. (1997).Implementing change from discipline-based teaching to interdisciplin-

ary teaming: A case study of five high school teachers. Unpublished doctoral dissertation, ThePennsylvania State University.

Murphy, Joseph (1991).Restructuring schools: Capturing and assessing the phenomena. New York:Basic Books.

National Center for Education Statistics (1994).Second follow-up: Student component data file user’smanual. Washington, DC: U.S. Department of Education, Office of Educational Research andImprovement.

Newmann, Fred M. (1992).A developing design. A leader’s guide to school restructuring: A spe-cial report of the NASSP Commission on restructuring. Reston, VA: National Association ofSecondary School Principals, pp. 30–36.

Newmann, Fred M. & Wehlage, Gary G. (1995).Successful school restructuring. Madison, WI:Center on Organization and Restructuring of Schools.

Perrow, Charles (1999).Normal accidents: Living with high risk technologies. Princeton, NJ:Princeton University Press.

Pogrow, Stanley (1997, November 12). The tyranny and folly of ideological progressiv-ism. Education Week(On-line article available at HYPERLINK http://www.edweek.org;http://www.edweek.org§/ew/1997/12pogrow.h17).

Rock, Donald A., Pollack, Judith M., & Quinn, Peggy (1995).Psychometric report for the NELS:88base year through second follow-up. Washington, DC: U.S. Department of Education, Office ofEducational Research and Improvement (NCES 95-382).

Shouse, Roger C. (1996a). Academic press and sense of community: Conflict, congruence, andimplications for student achievement.Social Psychology of Education, 1(1), 47–68.

Shouse, Roger C. (1996b). Academic press and sense of community: Conflict and congruence inAmerican high schools.Research in Sociology of Education and Socialization, 11, 173–202.

Shouse, Roger C. (1998). Restructuring’s impact on student achievement: Contrasts by schoolurbanicity.Educational Administration Quarterly, 34 (Supplemental), 677–699.

Biographical notes

Roger C. Shouseis an assistant professor of education at The Pennsylvania StateUniversity. A former secondary school teacher with over a dozen years experiencein both suburban and innercity settings, he received his Ph.D. in education fromThe University of Chicago. His research focuses on the diverse effects of schoolorganization, culture, and instructional practices.

Page 15: High risk, low return: The achievement effects of restructuring in disadvantaged schools

RESTRUCTURING IN DISADVANTAGED SCHOOLS 259

Lawrence J. Mussolineis the superintendent of schools in the Pine Grove AreaSchool District, Pine Grove, Pennsylvania. He received his Ph.D. from ThePennsylvania State University and has over 20 years of experience as a teacher,coach, and administrator in public and private schools.