Thrupp, M. (1999). Schools Making a Difference: Let's be Realistic! School Mix, School Effectiveness and the Social Limits of Reform. Reviewed by Bev Rogers, University of South Australia

Thrupp, M. (1999). Schools Making a Difference: Let's be Realistic! School Mix, School Effectiveness and the Social Limits of Reform. Buckingham, Philadelphia: Open University Press Pp 225 $45       ISBN 0-335-20212-8. Reviewed by Bev Rogers University of South Australia April 10, 2000         My background as principal of a school in a low socioeconomic area drew me to read this book, to explore the claims made of it by other writers. For example: Important research by Martin Thrupp (Thrupp, 1999) in New Zealand also points to how school effectiveness is a product of the social mix of students, and that the social mix of schools shapes teacher expectations, school organisation, curriculum and, therefore, student outcomes. That is, context factors outside a school's control have significant impact on student outcomes. (Blackmore, 1999, p. 35)         Based on my experiences, there is much in the above statement with which I can agree. Thrupp himself summarises the central theme of the book as "the 'schools can make a difference' message has been thoroughly overplayed" (Thrupp, 1999, p. 4). He is critical of the school effectiveness research (SER) and the emerging effectiveness and improvement (E&I;) work, judging these research traditions as insufficiently critical of the relationship between schools' intake and student achievement.         Thrupp uses his book to provide evidence that the claims of SER and E&I; work need to be qualified by the likely impact of "school mix," which he defines as, the social class composition of a school's student intake. According to Thrupp, evidence showing that school mix has a significant influence on school processes would challenge the current orthodoxy promulgated by researchers in the effectiveness and improvement traditions. Thrupp is critical of E&I; literature not just because the research is incomplete but also because it lends credibility to and provides a basis of support for the politics of polarisation and blame.         By "politics of polarisation," Thrupp means the increasing social class differentiation of school intakes as a result of parental "choice" in education as well as the growing disparities in educational resources and educational quality that result from this differentiation. "Politics of blame" refers to the construction of these disparities as simply a technical rather than a social or political matter by those who place faith in market forces. A neo-liberal strand of the "politics of blame" attributes the popularity of schools to their performance rather than their intakes. Thrupp claims that, because of their adherence to the view that schools do make a difference, the E&I; writers take little account of differences between rich and poor schools. According to Thrupp, governments have often used E&I; arguments to construct school failure as the responsibility of schools alone without any reference to the broader socio-political context, such as the impact of poverty. In such a frame, the claim that student achievement is significantly affected by socio-economic status and school mix is ruled out as a plausible excuse for poor performance.         The question of the impact of school mix on the performance of students goes to the heart of the contemporary debate over education policy. "The history of this concept over the past 30 years has been so coloured by political, ideological and methodological considerations that it is hard to establish what influence school mix has on student performance" (Thrupp, 1995, p. 183). Thrupp reveals that, whilst some quantitative research shows that school mix has a significant influence on school performance, there is a considerable body of qualitative research that makes this claim problematic.         Thrupp's review of the literature illustrates that the relationship between school mix and school achievement continues to confound school effectiveness research. He claims that school effectiveness researchers in the 1970's and 1980's responded to the problems of school inequities simply by denying such problems and arbitrarily asserting that schools could make a difference. These researchers reasoned that, if exemplary schools existed, then specific, identifiable, and reproducible characteristics of these schools would explain their success (Thrupp, 1995). According to Thrupp, these researchers deployed the concept of "ethos" or "climate"—a variable under the supposed control of schools—to explain the processes that resulted in school effectiveness. And they ignored the alternative explanation that "school mix"—a variable not easily amenable to control by school personnel—might influence the school conditions that lead to or constrain effectiveness (Thrupp, 1995; 1999).         In contrast to researchers in the school effectiveness and effectiveness and improvement traditions, Thrupp argues that the accumulated research on the topic does support the claim that there may be a significant relationship between "school mix" and student achievement. He acknowledges, however, that, in order to substantiate this claim, certain questions must be answered: (1) does a "school mix effect" exist or is it a proxy for another variable? (2) if a "school mix effect" does exist, how robust is it? (3) what produces the "school mix effect?"         As Thrupp acknowledges, finding the causes for phenomena like school achievement is tricky: "... quantitative research can find correlations but not investigate the causal processes that underlie them, while qualitative research can investigate those processes in more depth but not demonstrate causality either" (Thrupp, 1999, p. 43). Thrupp attempts to work around this difficulty by constructing a logical argument in answer to the question: "if there were a causal chain that explained school achievement, how might the variables in that chain be related?" Whilst I have no issue with his approach, I am troubled by the fact that later in the book, when he is drawing conclusions, Thrupp appears to forget about the tentative way in which he bracketed this causal claim. Within the brackets, Thrupp draws on a range of research literature to hypothesise causal mechanisms and how they might work. Thrupp lists three possible mechanisms that might together account for the "school mix effect": reference group processes, instructional processes, and organisational and management processes. He then conducts empirical research to substantiate the effect of these purported causes of school achievement.         Thrupp's research compares 13 matched pairs of students enrolled in four secondary schools. In selecting students to participate, Thrupp devised a method—based on Brown's Frame of Reference—to select working class students who were neither alienated from school nor particularly positive towards it. He refers to these students as "ordinary kids." Thrupp then created pairs matched on socioeconomic status, prior attainment, and orientation to school. The students differed with respect to the school they attended, but all of them were year-9 students just entering secondary school. Thrupp collected data using group interviews, student questionnaires, classroom and school observations, staff interviews, staff questionnaires, student work samples, teacher evaluations, other school documents, and analyses of the socio-economic characteristics of the schools. Using the Elley-Irving index (Elley & Irving, 1985), the schools were compared in terms of percentage of low, medium, and high SES students. On this basis, the schools were classified as working class, middle class, or somewhere between working and middle class. Thrupp's approach, however, seems to misconstrue what the Elley-Irving index purports to measure, as the following discussion of this index suggests: Socio-economic status is defined by the authors in terms of an equal weighting of the median educational and income levels for workers in each specific occupation group as reported in the 1981 Census. It is thus an objective index of occupational status and is descriptive not evaluative. It is not, and could not be, an index of social class, or opportunity to learn, or material well-being, or life-style, or consumption patterns or home circumstances. (p.116) If he had used the term "socio-economic status" in the classification, Thrupp might have better represented Elley and Irving's system of categorisation.         Despite some questionable methodological choices, Thrupp seems to capture many of the characteristic features of working class schools such as those in which I have worked. He identifies the day-to-day conditions that distinguish between schools where teaching is easy and schools where teaching is hard.         Thrupp exemplifies working class schools with two events that "ring true" for me. One event—students' lighting an aerosol can and using it as a blow torch—illustrates the "overwhelming" types of behaviour that occur in low-SES schools. The other event—a saga involving fees for exams—demonstrates that even well organised processes are thwarted by a lack of appreciation of deadlines and rules. Both of these events occurred at Tui College (the school classified as working class). Thrupp found that the "middle class" schools in the study could "get away with" a more laissez-faire style of organisation. Based on these examples, I cannot doubt Thrupp's claim that he collected adequate data of sufficient depth to differentiate the processes that occur in schools with different "school mixes."         What I do have difficulty with, however, are (1) Thrupp's assumption that the only difference between the schools was "school mix" and (2) his conclusion that the major differences in outcomes for the students in these schools resulted from differences in "school mix." Although Thrupp did attempt to control for differences in teachers, curriculum, and teaching methods at the different schools, he did not control for the schools' different levels of academic focus. This oversight constitutes an important problem because the data clearly suggest that academic focus was both openly and intentionally lacking at Tui College but was evident at the three other schools.         Given this circumstance, differences in student outcomes at the different schools might result from differences in "school ethos" rather than from differences in "school mix." And this possibility is clearly relevant to Thrupp's argument, which purports to demonstrate the flaws of the "school ethos" argument offered by E&I; researchers.         Thrupp comments that the middle class schools were better able than the working class schools to support academic programs. Moreover their students were more compliant and more able to cope with difficult work than were students in the working class schools. In the middle class schools, teachers assigned more demanding texts and made use of more academically focused teaching resources. In these schools, in fact, it seemed that school mix and "school ethos" were impossible to separate.         One of the most interesting and useful ideas in the book concerns the negotiation of curriculum and methodology that occurs as a result of students'"facilitation." In other words, students "facilitate" certain kinds of methodology and curriculum through their behaviour and, by means of this process, negotiate classroom pedagogy. According to Thrupp, the negotiation process takes place when teachers' behaviours are affected by those behaviours of students that mark their social class.         As a result of his analyses, Thrupp concludes that, while there was substantial within-school variation in both the academic difficulty of work and students' levels of engagement, the between-school variation was also notable. In particular Tui College classes were less academically demanding than classes at the middle class schools. And student engagement was also less evident at Tui College than elsewhere. Thrupp infers from this finding that the SES mix of the school significantly influences teacher behaviour, which in turn influences student achievement. Although this finding resonates with many teachers' experiences of different schools, it is not adequately explored in Thrupp's study. In particular, Thrupp neglects to explore the dynamics of the process. He ignores the question: "How does this actually happen in a low-SES school?"         Rather than addressing this question, Thrupp accepts the influence of school mix on teachers as a "given," and he goes on to make what, I think, is a sweeping claim: "the curriculum management and course offerings, classroom discipline, teaching approaches, curriculum content, assessment, teaching resources and teacher characteristics may all be diverse within schools but...this diversity is nevertheless bounded by school mix" (p. 81). Thrupp does not explain either who or what exactly is "bound by school mix" and how the process of imposing and responding to limits might work. By failing to address these dynamics Thrupp has begged the fundamental question asked by Rutter (1979): "are schools the way they are because of the children they enroll or do the children enrolled in a school behave the way they do because of school influences?"         In reading the book, I wondered about what was happening at Tui College—features of its environment that were not adequately examined in the text. For instance, Thrupp notes, "not much student work at Tui College was marked, perhaps because there was neither parental nor student pressure to do so" (p. 94). He also gives examples of other organizational features that distinguished Tui College from the three middle class schools, including lack of administrative support for classroom teachers. In categorizing these as "school mix" effects and not as "school practice" effects, Thrupp seems to be engaging in circular reasoning. Nevertheless, he uses this reasoning to support his argument for a "whole school effect," about which he concludes, "while schools may make some differences to student achievement, this is likely to be smaller than typically assumed by E&I; literature because of the way school mix affects school processes" (p.182).         In response to this conclusion, Thrupp repeats his argument about the social limits of school reform. He does, however, offer some alternatives—long-term solutions that he nevertheless suspects are unlikely to be implemented in the current political climate. These policy options include: (1) reducing the SES segregation of schools, (2) increasing resources substantially to low SES schools, (3) using fairer methods of accountability, (4) providing incentives for good teachers to move to low-SES schools, and (5) encouraging schools to use mixed ability grouping rather than tracking. Overall, Thrupp's argument moves from a position of caution concerning claims about causality to a depressing, almost fatalistic account of the way that school mix keeps schools from having an impact on the achievement of low-SES students. This reasoning, however, is based on just one empirical study. That study, moreover, is flawed. Whilst in many respects the study is thorough and uses a range of research techniques, it examines a very small sample of students and compares schools that differ in more ways than just their mix of students.         Rather than offering a mechanism for responding to the politics of polarisation and blame, I fear that Thrupp's fatalistic conclusions will lend support to policy arguments that relieve schools of the burden of providing high-quality education to low-SES students. After all, if such students remake schools in their own "image," they will invariably undermine schools' efforts to cultivate their achievement. This conclusion is extremely dangerous in an era in which school funding is tied to the "choices" made by individual (and mostly middle class) parents rather than to a socially responsible conception of the common good. References Blackmore, J. (1999). Framing the issues for educational redesign, learning networks a professional activism (Vol. 25). Hawthorn, Victoria: Publications Sub-Committee of the Australian Council for Educational Administration. Elley, W. B., & Irving, J. C. (1985). The Elley-Irving socio-economic index 1981 census revision. New Zealand Journal of Education Studies, 20(2), 115-128. Rutter, M., Maughan, B., Mortimer, P., & Ouston, J. (1979). Fifteen thousand hours. London: Open Books. Thrupp, M. (1995). The school mix effect: The history of an enduring problem in education research, policy and practice. British Journal of Sociology of Education, 16(2), 183-203. Thrupp, M. (1999). Schools making a difference: Let's be realistic! school mix, school effectiveness and the social limits of reform. Buckingham,Philadelphia: Open University Press. About the Reviewer Bev Rogers is a Principal of a large secondary school in Adelaide and an Ed.D student at University of South Australia (Underdale). Her research interests include educational disadvantage, vocational education and principal leadership.