Longitudinal Trends in Math and Science Partnership-Related Changes in Student Achievement with Management Information System Data Across Five Years (2003-04 to 2007-08)
Abstract
This substudy in the evaluation design of the Math and Science Partnership (MSP) Program Evaluation examines student proficiency in mathematics and science for the MSPs' schools in terms of changes across five years (2003/04, 2004/05, 2005/06, 2006/07, and 2007/08) and relationships with MSP-related variables using Management Information System data with the Annual K-12 District Survey. First, changes in percentages of students at or above proficient on state assessments in math and science were investigated by gender, ethnicity, special education, and students with limited English using the MIS data available for (a) across the five-year period (2003/04 - 2007/08) and (b) same schools across the last four years (2004/05-2007/08), with the purpose to obtain a sample of schools without missing data for dependable longitudinal analyses. The classification of MSP schools with and without focus on math or science for the longitudinal data over this four-year time period (2004/05-2007/08) was also taken into account. The results indicated that the MSP-related schools demonstrate sustained increase in percent of students at or above proficient in both math and science at all school levels. This trend was more clearly pronounced for schools with focus on math or science. Second, schools were examined by frequency and effect size of increase, decrease, or no change in student math and science proficiency. The schools with positive changes were in much higher numbers and higher mean effect size of change compared to schools with negative changes in student math and science proficiency. This trend was better pronounced for schools with focus on math at the elementary and middle school levels and for schools with focus on science also at the elementary and middle school levels. Third, longitudinal growth trajectories in mathematics and science proficiency across the four years (2004/05-2007/08) were investigated. The results indicated the existence of different latent classes of growth trajectories of school success on state assessments in mathematics and science
