Curriculum Change as a Reform Strategy: Lessons from the United States and Scotland
by Adam Gamoran - 1997
Recent research indicates that the school curriculum exerts an important influence on student learning: A rich and rigorous academic curriculum promotes high levels of student achievement, and curriculum differentiation is associated with achievement inequality. These findings suggest that curriculum change may be a potent policy lever. Two cases of planned curriculum change are examined to illustrate the limits and possibilities of curriculum reform. In the United States, many school districts are upgrading the quality of the mathematics curriculum for low-achieving students. Evidence from four urban districts shows that “transition courses?designed to bridge the gap between elementary and college-preparatory mathematics achieve partial suc -cess: Students in transition courses have better outcomes than those in general math, but are not as successful as those in college-preparatory classes. In Scotland, a national curriculum reform called “Standard Grade?was designed to enhance opportunities for disadvantaged students to study an academic curriculum in secondary school. Evidence from four longitudinal national surveys indicates that the reform raised achievement and reduced social inequality on national examinations at age sixteen, but inequality of enrolling in higher education persisted. These cases suggest that curriculum reform can have important benefits, but must occur in concert with other social changes to have broad and long-lasting effects.
Recent research indicates that the school curriculum exerts an important influence on student learning: A rich and rigorous academic curriculum promotes high levels of student achievement, and curriculum differentiation is associated with achievement inequality. These findings suggest that curriculum change may be a potent policy lever. Two cases of planned curriculum change are examined to illustrate the limits and possibilities of curriculum reform. In the United States, many school districts are upgrading the quality of the mathematics curriculum for low-achieving students. Evidence from four urban districts shows that transition courses designed to bridge the gap between elementary and college-preparatory mathematics achieve partial success: Students in transition courses have better outcomes than those in general math, but are not as successful as those in college-preparatory classes. In Scotland, a national curriculum reform called Standard Grade was designed to enhance opportunities for disadvantaged students to study an academic curriculum in secondary school. Evidence from four longitudinal national surveys indicates that the reform raised achievement and reduced social inequality on national examinations at age sixteen, but inequality of enrolling in higher education persisted. These cases suggest that curriculum reform can have important benefits, but must occur in concert with other social changes to have broad and long-lasting effects.
A growing body of sociological research shows that the manifest curriculum of schools plays a powerful role in influencing the outcomes of schooling. Two findings stand out: (1) A rigorous and meaningful academic curriculum enhances the productivity of schools, that is, students learn more when they experience a rich academic curriculum and; (2) differentiation in the curriculum leads to inequality of outcomes, that is, when students vary in their curricular experiences, they tend to differ from one another in what they learn (for reviews, see Gamoran, in press; Oakes, Gamoran, & Page, 1992). These findings have led some writers to conclude that access to rich curricular content is one of the most essential conditions for promoting student learning (e.g., Applebee, 1996; Porter, Archibald, & Tyree, 1991).
In light of the importance of curriculum for levels and distributions of student outcomes, it seems wise to consider curriculum reform as a strategy for improving the achievement of all students. What are the prospects for curriculum change as a reform strategy? By improving students access to high-quality curricular content, can educators produce higher levels of student learning with less inequality? In this article I consider the prospects for curriculum reform as an approach to educational change. I first offer a theoretical account of why modifying the curriculum may be a potent strategy. Next I assess two cases of curriculum change, which illuminate both the limits and the possibilities of curriculum reform. One case focuses on upgrading the mathematics curriculum for low-achieving students in low-income urban high schools in the United States. The second case examines a national curricular reform in Scotland, where secondary curricula and national examinations are closely aligned.
SOCIOLOGICAL PERSPECTIVES ON CURRICULUM
Early sociological studies of curriculum mainly considered the hidden curriculum, the subtle meanings conveyed to students through the structures and symbols of schooling, beneath the surface of the more obvious course of study (e.g., Apple, 1979; Whitty & Young, 1976; Young, 1971). Students everyday experiences in school were seen as operating to socialize students to accept their future positions in the larger social order. Thus, the hidden curriculum of schools played an important role in class reproduction (Bowles & Gintis, 1976).
Although most work in this literature was not empirical, a few important observational studies helped corroborate the theoretical claims. Keddie (1971) explored curricular differences in a streamed (tracked) school. In A-stream classes, students questions, which were generally consistent with the larger direction of the lesson, were treated as legitimate by teachers, who tried to respond. By contrast, students in C-stream classes who asked questions were seen as challenging the teachers authority or the lessons legitimacy, and the questions were typically dismissed by the teacher. Students social-class backgrounds tended to correspond to their streams, so students from advantaged social origins found their concerns validated by the schooling experience while disadvantaged students were pushed away. A few years later Metz (1978) reported a similar pattern in an American junior high school: High-track students questions were taken seriously, while low-track students questions were seen as disruptive.
While Keddie (1971) and Metz (1978) uncovered differences within schools in the hidden curriculum, Anyon (1981) found that schools in varied social contexts differ from one another in the way knowledge was presented to students. Based on fifth-grade social studies classes in five schools, Anyon showed that knowledge in schools located in affluent neighborhoods tended to emphasize creativity and original thought, whereas schools attended by working-class children more often presented knowledge as fragmented bits of basic facts.
Oakes (1985) continued this tradition of studying curriculum differentiation by concentrating on the actual content, materials, and activities of instruction in classes at different track levels. Her findings conveyed sharp contrasts between the serious academic content of high-track classes and the less rigorous content of instruction in low-track classes. Oakess study completed a transition in the study of curriculum from subtle and often hidden messages to manifest, overt content. However, as in earlier studies, Oakes had no measure of student learning, so the link between curriculum and achievement had yet to be established.
THE MANIFEST CURRICULUM AND STUDENT ACHIEVEMENT
A newer tradition in sociological studies of curriculum has uncovered clear connections between exposure to curriculum and student learning. Barr and Dreeben (1983) viewed the school as a system of nested layers, in which outputs at one level of school organization serve as inputs for the next. In this view curriculum is an element of the technology of instruction, allocated from administrators to teachers, who exert some degree of autonomy in providing access to curriculum to students. Barr and Dreeben showed that curriculum coverage in first-grade reading instruction exerts substantial effects on students learning to read. The more new words teachers introduced, the more words students learned to read, and the better they could read at the end of the year. Similarly, Rowan and Miracle (1983) found that teachers who covered more stories in the reading curriculum produced more learning among their elementary school students. Gamoran (1986) also showed that coverage of the reading curriculum largely accounted for the effects of ability grouping in first-grade reading.
Evidence for the importance of curriculum is equally strong at the secondary level. A variety of studies show that students who take more academic courses, and especially more advanced academic courses, gain more on achievement tests (e.g., Gamoran, 1987). These effects seem especially strong in mathematics and science, where achievement differences among students in different tracks is largely a consequence of differential course-taking. Bryk, Lee, and Holland (1993) attributed the achievement advantages of Catholic over public high schools to the more focused and rigorous academic curriculum of Catholic high schools. Consistent with this claim, Gamoran (1992) observed less inequality between tracks in Catholic compared with public high schools, presumably because Catholic high schools provide a demanding academic curriculum even to students outside the academic track.
In light of these findings, it is important to ask whether manipulating the curriculum could serve as a potent policy lever for raising student achievement. If particular emphasis is placed on increasing the richness and rigor of the curriculum available to low-achieving students, will that reduce the learning deficit suffered by such students over time? It may seem obvious that students cannot learn what they have not been taught, but this key point has failed to receive much attention in many plans for restructuring and reform (Murphy, 1991; Newmann & Associates, 1996).
CURRICULUM STRCUTURE AND CURRICULUM CHANGE
In contrast to many other countries, the curriculum in the United States is relatively undifferentiated (Rubinson, 1986). Almost all students study in comprehensive high schools, and there are relatively few formal divisions within schools. Nonetheless, differentiation occurs in American schools, ranging from reading groups in the early elementary years to varied course assignments in secondary schools. Although contemporary high schools rarely have formally identified tracking structures, the vast majority of schools divide students on some measure of ability for at least some subjects (National Center for Education Statistics, 1994).
When high schools sort students into honors, regular, and basic classes, curriculum and instruction are differentiated, and student achievement becomes correspondingly more unequal over time (Gamoran, Nystrand, Berends, & LePore, 1995). General or remedial mathematics classes, for example, do little to improve the achievement of students who are enrolled, while students in college-preparatory classes continue to learn (Gamoran, 1987; Kerckhoff, 1986). A fragmented and slow-paced curriculum is undoubtedly one of the reasons that students get so little out of taking general math (Oakes, 1985). One response to this problem is to upgrade the quality of the curriculum for low-achieving students. Instead of endlessly repeating the arithmetic curriculum of elementary school, some school districts now provide special transition courses designed to bridge the gap between basic and college-preparatory mathematics. The aim of transition courses is to bring low-achieving students into the college-preparatory pipeline. Can such courses succeed? In this article I report on an assessment of the effects of transition courses in four urban school districts in the United States.
Because curriculum differentiation leads to achievement inequality, one may suppose that standardizing the curriculum would provide greater support for equity. There are few if any models of large-scale curriculum standardization in the United States, and certainly none with a closely related examination system. To assess the impact of curriculum standardization, we need to look outside the United States. In Scotland during the 1980s, an important reform of the secondary curriculum, called Standard Grade, took place. Whereas students had previously been divided into academic and nonacademic courses of study, the Standard Grade reform brought all students into academic courses. What were the short-term and long-term results of this reform? As a second example of curriculum change, I will provide an assessment of the Standard Grade reform.
After examining the two cases of reform, I will discuss the broader prospects of curriculum change as a reform strategy. What are its limits and possibilities? What is the relation between the success of a curricular reform and the context in which it takes place? These questions are important for policy considerations.
UPGRADING HIGH SCHOOL MATHEMATICS FOR LOWACHIEVING STUDENTS
In response to the problems of lower-level mathematics classes described above, many states and districts are working to improve the quality of the mathematics curriculum for low-achieving students. The most recent and dramatic example of this effort is New York Citys decision to require all students to enroll in academic math and science courses in the first two years of high school. The New York City initiative is just getting under way, and assessment of the program is not yet possible, though early indications suggest that the program is being implemented (Diegmuller, 1995; Newman, 1995). Other districts have been enrolling low-achieving students in academic courses for several years, and a team of researchers at the Consortium for Policy Research in Education (CPRE) has gathered data on such programs from four districts: San Francisco and San Diego, California; and Rochester and Buffalo, New York. We selected California and New York because they had more extensive programs of upgrading lower-level math courses than other states. We focused on these four districts to emphasize urban locales with large numbers of low-achieving students, and programs that had been under way for several years. We aimed to select two high schools within each of the four districts, focusing on schools with high percentages of low-income and low-achieving students, and an upgraded math program in place for at least two years and involving at least three teachers. Only one school from Buffalo met these criteria, so only one was included, for a total of seven schools.
Each of these districts had implemented transition courses, designed to provide more meaningful and challenging mathematics to low-achieving students than had occurred in the general math classes, which were being replaced. The transition classes were supposed to facilitate students progression through a college-preparatory mathematics sequence. We studied three different types of transition classes:
In California, Math A was developed to replace high school general math classes with more rigorous and meaningful content, and to allow low-achieving students a chance to experience college-preparatory material. Math A was used in San Francisco and San Diego. Originally, Math A was supposed to replace the college-preparatory sequence, but in practice, it has more often been used as a bridge to the traditional college-preparatory sequence of algebra and geometry. Thus, Math As success would be indicated by large numbers of students who begin in Math A eventually managing to complete some college-preparatory courses. Math A involved changes in both content and pedagogy. The course integrates content from measurement, geometry, algebra, logic, and probability. It emphasizes problem solving, reasoning, and real-world application more than memorization and drill.
In Rochester, New York, teachers developed a sequence of courses that covered the same material as the college-preparatory Regents classes, but at a slower pace. Two years of Stretch Regents were equivalent to one year of the regular Regents sequence. Thus, over the four years of high school, a student enrolled in Stretch Regents could complete the equivalent of Regents I and II, the minimal requirement for many New York state colleges and universities. Formerly, a low-achieving student who entered high school in general math would subsequently enroll in business or consumer math, and never reach the college-preparatory curriculum. Stretch Regents aimed to change that pattern. The two Rochester high schools differed in the ways they organized Stretch Regents. In one school, teachers covered the regular Regents material at a slower pace, completing half the textbook each year. In the other school, teachers covered the whole textbook in one year, emphasizing only the easy parts, and went over the same textbook the next year, focusing this time on the hard parts.
In Buffalo, courses from the University of Chicago School Mathematics Project (UCSMP) served as an alternative to the Regents math sequence. UCSMP was adopted to counter the high failure rates in Regents courses. The first course in the UCSMP sequence, called Transition Math, is analogous to the other courses we studied, in that it provides a bridge between elementary and college-preparatory mathematics. UCSMP stresses an integrated mathematics content, including applied arithmetic, algebra, geometry, logic, probability, and statistics. Like Math A, the course emphasizes problem solving and real-world applications. A sign of success for UCSMP Transition Math would be high proportions of students pursuing a college-preparatory curriculum.
We asked two main questions about the effects of transition courses:
1. Did students who entered transition courses at the beginning of high school successfully navigate a college-preparatory math program by the end of high school? If so, this pattern would contrast with general math, which is typically a dead end for students.
2. Do students learn more meaningful mathematics in transition classes? How does their learning compare with what usually occurs in general math classes, on the one hand, and college-preparatory classes, on the other?
To address the first question, we examined the high school transcripts of a cohort of students in each of the four districts. To consider the second question, we administered tests of learning in the fall, winter, and spring of the 1992-1993 school year, gave questionnaires to students and teachers, and carried out observations of classroom instruction. White (1995) provides more information about the implementation of the transition courses, and White, Gamoran, Smithson, and Porter (1996); White, Porter, Gamoran, and Smithson (1996); and Gamoran, Porter, Smithson, and White, (1996) provide more information about the effects of the transition courses. The following results are drawn from those reports.
In comparing students enrolled in general, transition, and college-preparatory math classes, we noted that formal decision rules about assignment criteria were frequently not observed. For example, in one district, students who earned a C+ in eighth-grade math were supposed to be assigned to the transition course in ninth grade. In practice, less than 40 percent of such students were found in transition courses, and over 50 percent of such students were assigned to general or college-preparatory courses (White, Gamoran, et al., 1996). As DeLany (1991) has shown, logistical considerations often override educational criteria when students are assigned to classes. Although this pattern may be problematic for students, it is useful for researchers, because it weakens the correlation between students preexisting characteristics and their class assignments, making it easier to detect the effect of class assignment. Still, students enrolled in difference courses varied in their math performance prior to high school entry, so we attempted to take differential selection into account. In the transcript study, we first examined raw course-taking patterns, but in two districts we followed this with analyses that controlled for students math performance prior to entering high school. In the achievement study, we monitored student achievement growth over time, and controlled for academic and social background characteristics of students.
EFFECTS OF TRANSITION COURSES ON MATHEMATICS COURSE-TAKING
Four years of high school transcripts were drawn for over 1,400 students. (This figure excludes Buffalo because only three years of data were available there, since the transition course had been in place for three years at the time the transcripts were collected.) Attrition of students reduced the sample to fewer than 900 for analysis. Did students who started out in transition courses succeed in completing a college-preparatory curriculum? We defined a minimal college-preparatory sequence as completion of at least algebra and geometry (or equivalent). Figure 1 shows that students who enrolled in transition courses at the beginning of high school were substantially more likely to complete a college-preparatory sequence than students who started out in general math. In Rochester, for example, students who began in general math almost never made it through a college-preparatory sequence (only 1.6 percent did so). In contrast, 12.7 percent of students who started out in the transition class completed the minimal college-preparatory sequence. In San Francisco and San Diego, completion rates were higher overall, and the advantage of the transition course over general math is again evident. At the same time, completion rates in all three districts were highest for students who entered college-preparatory classes at the beginning of high school. Over 80 percent of students who entered college-preparatory classes in ninth grade successfully completed at least two years of a college-preparatory sequence.
We corroborated these results with several additional analyses. First, we obtained information from Rochester and San Diego on students prior math performance, to test whether the differences between general math and transition courses were statistically significant, controlling for prior performance. The advantage of transition courses was sustained. Second, we examined whether students accumulated math credits in the different types of courses, fearing that if transition courses were more challenging than general math, they might increase failure rates and prevent students from accumulating enough credits to graduate. This concern proved unfounded: Students in transition courses accumulated no fewer credits than students in general math. Third, we examined course sequences from year to year, to incorporate students who had been lost from our sample due to attrition or because we had only three years of data (i.e., Buffalo). We observed the same patterns from year to year and over three years as we had in the four years of data. These findings are reported in more detail by White, Gamoran, et al. (1996).
EFFECTS OF TRANSITION COURSES ON MATHEMATICS ACHIEVEMENT
We collected achievement tests from almost 1,100 students in the seven schools in the fall of 1992, with a response rate of 80 percent. Students were retested in the winter and spring, and a total of 882 students responded to at least two test administrations. We obtained information about student backgrounds from teacher and student questionnaires. Data on classroom instruction came from teacher questionnaires and classroom observations. Our analysis included forty-eight classes in the seven schools. (Originally, fifty-six classes were included, but six bilingual classes and two others were identified as outliers and were dropped from the analysis.)
The analysis used a multilevel statistical model of three time points for achievement nested within students, who were themselves nested within classes, and the model controlled for students sex, race, ethnicity, and prior math grade, as well as a class-level indicator of socioeconomic status. This analysis models growth in achievement at three levels. First, it estimates growth for each student, simply as a function of time. Because there were only three time points, growth was assumed to be linear. Second, the analysis examines differences among students in their rates of achievement growth, as a function of varying academic and social background characteristics. Third, the analysis estimates differences between classes in average achievement growth of students (adjusted for variation in student background characteristics) as a function of the different class types.
The left side of Figure 2 shows achievement growth in the different types of classes. (Due to small numbers of classes, Math A and UCSMP classes were included in the same category for this analysis. Also, the general math category includes pre-algebra classes.) Achievement growth in Regents classes was significantly greater than that of general math classes, consistent with previous research. Achievement in the transition classes fell in between that of the Regents and general math classes.
Were these achievement differences a result of varied curriculum content? We constructed an indicator of content coverage using our achievement test as a benchmark. We compared topics and cognitive demands made on the test with those teachers reported making in class. We found the most content coverage in Regents and algebra classes and the least in general math, with the transition classes again in between. We added content coverage to the statistical model and found that it had significant effects on achievement growth. Using the results of this regression, we simulated what achievement would have been like if all classes had covered the same curriculum. For example, if content coverage were as extensive in general math as it was in Regents, achievement growth in general math would be predicted to rise in accordance with the added content coverage. The right side of Figure 2 suggests that if the curriculum were fully standardized, achievement would be nearly equal among classes.
TAKING STOCK OF THE UPGRADED MATH CURRICULUM
The transition classes appear at least partially successful in fulfilling their aims. They improve students chances of completing a college-preparatory curriculum and probably benefit students learning as well. Certainly for students who move from transition to college-preparatory classes, learning improves. However, other things being equal, a given student will do best of all to start his or her high school career not in a transition class, but in a college-preparatory class in the first place.
Our findings imply that standardizing the curriculum would reduce inequality of achievement. The greater richness and rigor of the transition classes compared with general math probably contributes to their success. However this study, like most other studies of curriculum differentiation, examines a system at one point in time. If the system were to change, would the results really change as the model implies? For the United States, a few case studies indicate that curriculum change has been difficult, and its effects have been mixed (Wells & Sirna, 1996; Wilson & Rossman, 1993). No study in the United States has examined planned curriculum change on a national scale, and no broad study has used outcomes linked to the curriculum to examine the consequences of change, in part because such coordinated changes have not occurred on a wide scale. In other countries, however, wide-ranging changes in curriculum and examinations are not uncommon. One such case is Scotland during the 1980s, and for this case, a series of surveys is available to provide evidence about the impact of the change. We now turn to the Scottish case as an example of large-scale curriculum change over time.
CURRICULUM STANDARDIZATION IN SCOTTISH SECONDARY SCHOOLS
Until the early 1980s, Scottish secondary students were selected for academic study in the Ordinary Grade (O Grade) courses. Students who were not selected for O Grade courses could not take the corresponding examinations, and thus had no chance for further academic study in secondary school, let alone higher education. In an effort to broaden access to academic study, the Scottish education system implemented the Standard Grade reform beginning in the mid-1980s. Unlike the O Grade system, in which students were selectively enrolled in academic courses, the Standard Grade reform encouraged all students to take academic courses and to sit for the corresponding national examinations about age sixteen. Standard Grade courses were not totally undifferentiatedsome grouping occurred into levels called foundation, general, and creditbut these levels were mixed in most schools, most students sat for two levels of the examinations (i.e., foundation and general, or general and credit), and the system was far less differentiated than the academic versus nonacademic divide that had characterized the O Grade system (see Gamoran, 1996, for further details).
How did the Standard Grade reform affect student achievement and attainment? To address this question, I analyzed the Scottish Young Peoples Surveys (SYPS) for four cohorts of students who completed their last years of compulsory schooling (equivalent to grade ten in the United States) in 1984, 1986, 1988, and 1990. The four surveys were administered about nine months after the completion of compulsory schooling. At that point, some students were continuing in secondary school, others had entered nonacademic further education, and others were in jobs, training schemes, or unemployed. Students reported what they were doing, and also gave information on their O Grade or Standard Grade examination results, as well as family background data. SYPS carried out follow-up surveys of each cohort about two years after the original surveys, when students were about nineteen years old, and the follow-up surveys can be used to indicate whether students had entered postsecondary education by that time. Response rates to the first surveys averaged close to 80 percent, and about two-thirds of the initial respondents also participated in the follow-up surveys.
I used multilevel statistical models of students within schools to examine the effects of the Standard Grade reform. I compared schools in 1984, before the reform, with schools in 1990, after the reform, in English, mathematics, and science, the first subjects to be implemented for Standard Grade. For 1986 and 1988, I also compared schools that had implemented Standard Grade with those that had not yet done so. I examined whether achievement rose or fell during those years, and whether examination scores became more equally distributed by social background within schools, as a result of the reform. (These results were reported in Gamoran, 1996.) Finally, I used the follow-up surveys to examine whether Standard Grade had any impact on levels and distributions of continuing in secondary school and pursuing postsecondary education. (These results are reported here for the first time.)
EFFECTS OF STANDARD GRADE ON EXAMINATION RESULTS
On average, controlling for gender, social background, and school context, examination scores rose in English, mathematics, and the sciences during the period of the implementation of Standard Grade in these subjects (Gamoran, 1996). This rise was due in part to the reform itself: Scores rose faster in schools that implemented Standard Grade first. Moreover, the unequal distribution of examination results diminished over the same period. Within years, social inequality of achievement within schools was less in schools that implemented Standard Grade, compared with schools that had not. (Most of the within-year contrasts are not statistically significant, but the pattern is highly consistent across the years and subjects.)
Figure 3 illustrates the impact of Standard Grade on inequality in mathematics attainment. It examines the chances of receiving an award in mathematics, as opposed to no award at all (which would mean the student either did not take or completely failed the mathematics examination). This comparison is important because receiving an award both signifies access to the academic curriculum and indicates national certification in recognition of the students academic study. Two hypothetical students are compared: an advantaged student whose parents were highly educated professionals, and a disadvantaged student whose parents left school before age sixteen and held unskilled jobs. In 1984, the advantaged student had more than an 80 percent likelihood of receiving an examination result at some level, whereas the disadvantaged student had only a 20 percent chance of receiving any award at all. By 1990, the situation had changed dramatically; while the advantaged students chances rose to over 90 percent, the disadvantaged students opportunity rose even faster, to nearly 70 percent. Thus, socioeconomic inequality within schools in the opportunity for academic awards narrowed over the period of the reform. Additional analyses show similar results for science and English, although the change was not quite as dramatic in those subjects (see Gamoran, 1996). Moreover, the same pattern occurs within years for the comparison of schools that implemented Standard Grade with those that had not yet done so. (Again, the pattern is highly consistent, although most of the coefficients for within-year contrasts in social background effects are not statistically significant.)
Inequality within schools declined because disadvantaged students had better opportunities to enroll in academic courses, and they usually obtained awards in the subjects they studied. Further analysis indicates that most of the change occurred toward the lower end of the achievement scale (Gamoran, 1996). Figure 4 displays the odds of receiving a high mark, called a pass, of 1, 2, or 3 on the 7-point examination scale. This comparison is also important because high marks improve a students chances of further academic study after compulsory schooling. These odds changed much less over time, and disadvantaged students did not progress at a faster rate than their advantaged peers. (See Gamoran, 1996, for further discussion.)
EFFECTS OF STANDARD GRADE ON POSTCOMPULSORY EDUCATIONAL ATTAINMENT
Did Standard Grade change the levels and distributions of postcompulsory schooling? Figure 5 shows that the most privileged students almost universally continued their schooling in 1984 as well as in 1990. Staying-on rates rose for disadvantaged students, as seen in Figure 5, so that inequality declined slightly during this time period. Overall, rates of continuation beyond compulsory schooling increased, and this trend was probably caused at least in part by the Standard Grade reform, because in 1988, when some schools had implemented the reform more extensively than others, staying-on rates were significantly higher in schools that had adopted more Standard Grade courses. However, even though the reform led to a general increase in staying on beyond compulsory schooling, it cannot be cited as the cause of the reduction in inequality that occurred, because in 1986 and 1988, schools exhibited the same amount of inequality regardless of how many courses had been implemented for Standard Grade. Thus, the Standard Grade reform led to an increase in rates of staying on, but the small decrease in inequality must be attributed to other sources.
Finally, Figure 6 shows that although rates of entering postsecondary education by age nineteen rose over the time period, the increase was the same for advantaged and disadvantaged students, so there was no change in inequality. Within years, a schools adoption of Standard Grade was unrelated to both the level and the social distribution of entering higher education. I examined alternative indicators of postsecondary educationwhether the respondent was in postsecondary education at age nineteen, and restricting the definition to higher education (i.e., excluding vocational and technical postsecondary education)and obtained comparable results. Beyond secondary school, I did not find effects of the Standard Grade reform.
TAKING STOCK OF THE STANDARD GRADE REFORM
At one level, the Standard Grade reform has been dramatically successful. More students than ever are studying academic curricula and are obtaining awards on the national examinations. There is also less inequality within schools in who participates in academic study. These changes are, in part, a direct result of the reform. If we think about the education system as a ladder, then the Standard Grade reform allowed many more students to step onto the ladder. It has not, however, taken them all to the top of the ladder. Inequality remains at the highest levels of schooling.
There are some signs that inequality in access to postsecondary education may be diminishing. Paterson (1997), using data through 1994, observed a decline in social inequality of students going directly from secondary to higher education. This trend may be an indirect result of the Standard Grade reform, if the broader accumulation of academic qualifications has created greater demand for access to higher education, but it is difficult to confirm this speculation empirically. Whether or not Standard Grade has played a role in postsecondary expansion, it has clearly opened up the possibility of high levels of educational attainment for students who would formerly have been excluded.
Irrespective of its role in postsecondary education, the impact of Standard Grade at the secondary level, both for examinations and for staying on, is significant for several reasons. First, academic study has value in its own right, stimulating students intellectual development. Second, receiving awards on national examinations, even if they are not at the top levels, helps students in a wide range of subsequent activities, including employment and training. Third, Standard Grade removed the stigma of nonacademic study from the last years of compulsory schooling, as reforms currently underway in Scotland are attempting to accomplish for postcompulsory secondary schooling (Scottish Office Education Department, 1994).
THE LIMITS AND POSS IBILITIES OF CURRICULUM REFORM
The cases I have examined here confirm the claims of previous research. Clearly, curriculum has an impact on student learning. Moreover, curriculum can be manipulated. Consequently, changes in curriculumacross classes, across schools, and over timelead to changes in student achievement. These findings are consistent with Barr and Dreebens (1983) organizational conception of schools, in which curriculum is a technological tool wielded by teachers to stimulate student learning. Policymakers seeking leverage over learning would do well to consider curriculum change as a path to improvement.
Indeed, curriculum change has been under consideration for some time in U.S. policy circles. Standards for curriculum and assessment are widely discussed at the national, state, and local levels across the country. Beyond increased high school course-work requirements (which, according to Wilson and Rossman , often led to increases in enrollment in low-level academic courses, precisely those that this study suggests should be avoided), little actual headway has been made in reforming curricula on a large scale. Despite the symbolic power of the federal Goals 2000 initiative (Borman, Cookson, Sadovnik, & Spade, 1996), and the innovative platforms of subject matter groups (e.g., National Council of Teachers of Mathematics, 1989), serious implementation of curricular reforms has been limited to local initiatives at best.
A major problem for curriculum reform in this country is the absence of widely recognized examinations that are tied to the curriculum. In countries with articulated systems of curriculum and examinations, such as Scotland, the curriculum is a serious entity, not to be trifled with or ignored. In the United States, only normative pressures (and loose state standards) sustain some degree of curricular coherence from one locale to the next. Curriculum reforms without assessments are likely to be weakly implemented and thus to have modest effects. As Smith and ODay (1991) argued, systemic reform calls for an alignment of curriculum and assessment. Statewide assessments, for example, would make state curricular reforms much more powerful than commonly occurs at present. In this sense, the context of curriculum reform is likely to make a substantial difference for the effects of curriculum reform.
One concern about combining curriculum reform with new assessments is that teachers would teach to the test. That, of course, is the point. If the test is worthy, then teaching to it is an appropriate goal. That would be the case if curriculum and assessment were aligned. This conclusion also indicates that appropriate teacher preparation and opportunities for professional development are necessary ingredients of systemic reform.
Another concern is that a high-stakes examination systemone that has real consequences for students futures, and perhaps for those of educators as wellmay uncover new inequalities among students, particularly emphasizing racial and ethnic differences in examination performance. While this concern is important, I think it most likely that upgrading the curriculum to meet new assessment standards would benefit low-income and minority students most of all, because their present curricula are weak, as noted in the math-upgrading study. Both that study and the assessment of the Standard Grade reform suggest that disadvantaged students can benefit from curriculum change.
The two cases I studied also reveal some limits of curriculum reform. Both investigations indicate that reducing differentiation in the curriculum tends to lessen inequality, but only to a point. In the upgrading study, low-achieving students came closer to their college-preparatory peers, but did not reach equality. In the Standard Grade analysis, access and certification improved for disadvantaged students, but the benefits were limited to secondary education. These limits make sense: Educational reforms have their largest effects on the most proximate outcomes. Broader changes are much more complicated, because they are affected by a much wider range of causes. To improve equity in access to higher education in Scotland, for example, it is no doubt necessary to affect not only the early secondary curriculum, but the later secondary curriculum and perhaps the structure of higher education as well.
What are the prospects for curriculum reform in American education? In one sense they have never been better. National (largely symbolic) goals emphasize raising standards; federal funds are available, through the Goals 2000 program, to help states and districts establish standards for themselves; teacher professional development has been added to the list of national goals; and subject-matter groups are helping to develop the potential content of new curriculum standards. However, unless curriculum change is accompanied by new assessments, for example at the state level, and linked to teacher preparation and professional development, the impact of curriculum reform will undoubtedly be modest at best.
An earlier version of this paper was presented at the Conference on Governance and Reform sponsored by the Sociology of Education Section of the American Sociological Association and the Teachers College Record, New York City, August 21, 1996. The research was supported by a grant from the National Science Foundation (REC-9627616), and by the Office of Educational Research and Improvement through grants to the Consortium for Policy Research in Education (R117-G10007) and the National Center for Improving Student Learning and Achievement in Mathematics and Science (R305-A60007). The author is grateful for advice and assistance from fellow researchers, programmers, and support staff at the Centre for Educational Sociology, University of Edinburgh, and the Wisconsin Center for Education Research, University of Wisconsin-Madison. The editor and reviewers of Teachers College Record provided helpful comments on an earlier version of the paper. Findings and opinions are those of the author and do not necessarily represent the views of the sponsoring agencies.
Anyon, J. (1981). Elementary schooling and distinctions of social class. Interchange, 12, 118-132.
Apple, M. W. (1979). Ideology and curriculum. London: Routledge & Kegan Paul.
Applebee, A. N. (1996). Curriculum as conversation: Transforming traditions of teaching and learning. Chicago: University of Chicago Press.
Barr, R., and Dreeben, R. (1983). How schools work. Chicago: University of Chicago Press.
Borman, K., Cookson, P. W., Jr., Sadovnik, A., & Spade, J. Z. (Eds.). (1996). Implementing federal legislation: Sociological perspectives on policy. Norwood, NJ: Ablex.
Bowles, S., & Gintis, H. (1976). Schooling in capitalist America. New York: Basic Books.
Bryk, A. S., Lee, V. E., & Holland, P. B. (1993). Catholic schools and the common good. Cambridge: Harvard University Press.
DeLany, B. (1991). Allocation, choice, and stratification within high schools: How the sorting machine copes. American Journal of Education, 99, 181-207.
Diegmuller, K. (1995, May 17). NYC requirements seen spurring gains. Education Week, p. 8.
Gamoran, A. (1986). Instructional and institutional effects of ability grouping. Sociology of Education, 59, 185-198.
Gamoran, A. (1987). The stratification of high school learning opportunities. Sociology of Education, 60, 135-155.
Gamoran, A. (1992). The variable effects of high school tracking. American Sociological Review, 57, 812-828.
Gamoran, A. (1996). Curriculum standardization and equality of opportunity in Scottish secondary education, 1984-1990. Sociology of Education, 29, 1-21.
Gamoran, A. (in press.). In D. Levinson (Ed.), Education and sociology: An encyclopedia. New York: Garland
Gamoran, A., Nystrand, M., Berends, M., & LePore, P. C. (1995). An organizational analysis of the effects of ability grouping. American Educational Research Journal, 32, 687-715.
Gamoran, A., Porter, A. C., Smithson, J., & White. P. A. (1996). Upgrading high school math instruction: Improving learning opportunities for low-achieving, low-income youth. Madison, WI: Consortium for Policy Research in Education.
Keddie, N. (1971). Classroom knowledge. In M. F. D. Young (Ed.), Knowledge and control (pp. 133-160). London: Collier-Macmillan.
Kerckhoff, A. C. (1986). Effects of ability grouping in British secondary schools. American Sociological Review, 51, 842-858.
Metz, M. H. (1978). Classrooms and corridors: The crisis of authority in desegregated secondary schools. Berkeley: University of California Press.
Murphy, J. (1991). Restructuring schools: Capturing and assessing the phenomena. New York: Teachers College Press.
National Center for Education Statistics. (1994). Curricular differentiation in public high schools (Document No. NCES 95-360). Washington, DC: U.S. Department of Education.
National Council of Teachers of Mathematics. (1989). Curriculum and evaluation standards for school mathematics. Reston, VA: Author.
Newman, M. (1995, August 31). Shaky 9th grade test results as math and science toughen. New York Times, p. A16.
Newmann, F. M., & Associates. (1996). Authentic achievement: Restructuring schools for intellectual quality. San Francisco: Jossey-Bass.
Oakes, J. (1985). Keeping track: How schools structure inequality. New Haven: Yale University Press.
Oakes, J., Gamoran, A., & Page, R. N. (1992). Curriculum differentiation: Opportunities, outcomes, and meanings. In P. W. Jackson (Ed.), Handbook of research on curriculum (pp. 570-608). New York: Macmillan.
Paterson, L. (1997). Trends in higher education participation in Scotland. Higher Education Quarterly, 51, 29-48.
Porter, A. C., Archbald, D., & Tyree, A. (1991). Reforming the curriculum: Will empowerment policies replace control? In S. H. Fuhrman & B. Malen (Eds.), The politics of curriculum and testing: 1990 yearbook of the Politics of Education Association (pp. 11-36). London: Taylor and Francis.
Rowan, B., & Miracle, A. W., Jr. (1983). Systems of ability grouping and the stratification of achievement in elementary schools. Sociology of Education, 56, 133-144.
Rubinson, R. (1986). Class formation, politics, and institutions: Schooling in the United States. American Journal of Sociology, 92, 519-548.
Scottish Office Education Department. (1994). Higher still: Opportunity for all. Edinburgh: Author.
Smith, M., and ODay, J. (1991). Systemic school reform. In S. H. Fuhrman & B. Malen (Eds.), The politics of curriculum and testing: 1990 yearbook of the Politics of Education Association (pp. 233-267). London: Taylor and Francis.
Wells, A. S., & Sirna, I. (1996). The politics of culture: Understanding local political resistance to detracking in racially mixed schools. Harvard Educational Review, 66, 93-118.
White, P. A. (1995). Math innovations and classroom practice: Upgrading the math curriculum at the high school level. Madison, WI: Consortium for Policy Research in Education.
White, P. A., Gamoran, A., Smithson, J., and Porter, A. C. (1996). Upgrading the high school mathematics curriculum: Math course-taking patterns in seven high schools in California and New York. Educational Evaluation and Policy Analysis, 18, 285-307.
White, P. A., Porter, A. C., Gamoran, A., & Smithson, J. (1996). CPRE Policy Brief: Upgrading high school math: A look at three transition courses. Philadelphia: Consortium for Policy Research in Education.
Whitty, G., & Young, M. F. D.(Eds.). (1976). Explorations in the politics of school knowledge. Driffield, England: Studies in Education Ltd.
Wilson, B. L., & Rossman, G. B. (1993). Mandating academic excellence. New York: Teachers College Press.
Young, Michael F. D. (Ed.). (1971). Knowledge and control. London: Collier-Macmillan.