|General Info | Resources | Research & Reports | Contacts & Connections | Curriculum|
Standardized Tests: Highlights from Current Studies of IMP® Student Performance
During 199495, the IMP curriculum was taught in 99 high schools in 12 states across the country (AZ, CA, CO, CT, HI, IL, MA, MN, OR, PA, TX, WA). Many of these sites are engaged in evaluation procedures for their own districts. We have received several reports from the sites concerning IMP student achievement on standardized tests. We also have preliminary reports from the WCER study that are related to standardized test achievement. These reports show that IMP students are doing as well as, and in some cases better than, students in traditional mathematics classes.
Philadelphia Regional Data
Central High School is a magnet school in Philadelphia, where IMP has been taught since 1993. The student body is selected on the basis of superior academic records and rigorous entrance examinations. A study comparing tenth-grade IMP students with non-IMP students at Central High showed IMP students scoring significantly higher, not only on the math component of the PSAT test (483 vs. 463) but also on the verbal test (514 vs. 487). Both results were significant at the .01 level. The student sample in this study was composed of all tenth grade students at Central High who were enrolled either in Year 2 of the IMP program or in a traditional geometry course (including those in Honors programs).
Denver Regional Data
In Denver, where IMP has been taught since 1990 at Eaglecrest High School, students took the math portion of sample SAT tests at the beginning and at the end of the 199192 academic year (Table I). The average gain of IMP Year 1 students was significantly higher than that of algebra I students. Average raw scores of IMP Year 1 students increased by 2.92, while average raw scores of algebra I students increased by only 1.25. The difference in growth was significant at the .025 level.
California Pilot Study
During the spring of 1992, when IMP was in its third year of implementation at its three original test sites, a sample SAT was administered to 285 students from IMP and control classes. An NSF impact study examined the results and observed that IMP scores were just as high as those of the control group, and added that IMP "has made significant progress in developing and field testing an integrated, four-year, problem-based mathematics curriculum for secondary school students."*
WCER Evaluation: School No. 2 Matched Groups
One aspect of the Wisconsin Center for Education Research (WCER) evaluation of IMP involves a comparison of two carefully matched groups of students at a California high school ("School No. 2") where the curriculum has been taught since 1989 (Table II). This comparison is part of the Class of 1993 Transcript Study. The IMP group (69 students) consists of all students who enrolled in the Interactive Mathematics Program in ninth grade who did not transfer out of school or into another mathematics program during their high school career. The non-IMP group (72 students) consists of students who were enrolled in traditional mathematics course offerings since ninth grade. The non-IMP group was selected to match the IMP group, by mean score and standard deviation, on a prehigh school measure of mathematics achievement (seventh grade scores on the CTBSComprehensive Tests of Basic Skills). The two groups were also matched to reflect gender and ethnic balance. The IMP group achieved a significantly higher overall grade point average than the non-IMP group and took significantly more semesters of mathematics than the non-IMP group. Although the data shows SAT scores of the two groups to be virtually identical, it should be noted that 87% of IMP students in the group took the SAT while only 58% of the non-IMP students took the test. Thus, the IMP group maintained a comparable level of achievement, while expanding the pool of students interested in meeting college entrance requirements.
From our perspective, the observation that IMP students are holding their own on standardized tests is especially meaningful because we know that these students devote perhaps 2030% of their classroom time over the four-year span of the curriculum to learning important topics that are not covered in traditional curricula. These topics include statistical reasoning, use of matrices, and fundamentals of calculus.
Standardized tests do not incorporate this mathematical content, nor do they examine other key aspects of IMP, such as the ability to work in groups or to write substantial analyses of complex mathematical problems. Throughout the country, education research groups are developing new tools that would shift the emphasis away from timed tests of isolated skills toward a more detailed assessment of mathematical understanding. The IMP evaluation team supports this effort, and is working to identify alternative assessment tools that will adequately measure IMP students' success in mathematics.
Schoen, Harold, Norman Webb, and Sharon Whitehurst. Dissemination of Nine Precollege Mathematics Instructional Materials Projects Funded by the National Science Foundation, 198191, April 1993, Wisconsin Center for Education Research, School of Education, University of Wisconsin-Madison.