The purpose of this proposal is to request funds to examine in depth data collected during the past three school years (1997 - 2000) for the "Longitudinal/Cross-Sectional Study of the Impact of 'Mathematics in Context' on Student Performance," funded by the National Science Foundation. This study is an example of a research study on student learning of mathematics in complex educational contexts.
The research model for the study is a structural model for monitoring changes in school mathematics designed to capture the impact that the dynamic, complex interaction of social context for learning, curricular content and materials, classroom experiments, and pupil motiviation have on student performance and further pursuit of mathematics. The mo9del is composed of 14 variables and their theoretical interrelationships. The study also featured a nonequivalent control-group, quasi-experimental design. about 80% of the students and teachers used "Mathematics in Context"; the remaining 20% of the students used conventional curricla already available in the schools. The longitudinal parat of the study involves tracking the performance of the three cohorts of students over the three years of study, and the cross-sectional part includes three cross-grade comparisons by year and three cross-year comparisons by grade. Information from teachers, administrators, and students in grades 5, 6, 7, and 8 in four school districts has been collected and is now being entered into a database.
The analyses will be directed toward both the longitudinal and cross-sectional comparisions an dtoward a set of questions about the implementation of standards-based programs, such as: "Do students continue to develop number skills in sucha a program?" or "Are such programs effectivie with students with differing backgrounds?" or "Do teachers change their pedagogy when presenting content in non-traditional ways?" To make the comparisons and to answer such questions the analyses will include exploratory descriptions of information related to each variable in the model, linear regression for a simplified model, structural equations to estimate path coefficients in the complete model, and a series of sub-studies examine in depth some patterns of relationships in the data. Finally, based on these analyses a series of reports are palnned. Currently a set of technical reports are being written on the creation of the data consturction instruments, the indicies for each variable, and the raw data that has been collected. Funds in this proposal would make it possible to report on the cross-sectional and longitudinal findings and our answers to the set of questions about the "Mathematics in Context" (National Center for Research in Mathematical Sciences Education & Freudenthal Institute, 1997-98).
