| This study investigated the effects of curricular tracking in high school mathematics on student achievement. Psychometric methods used to investigate item and test bias are applied to student groups that are distinguished by differences in intended and implemented curriculum. Specifically, an investigation of differential item functioning (DIF) and differential bundle functioning (DBF) on the tenth-grade Illinois Goal Assessment Program test items and item bundles documents the attenuating effects of tracking on student achievement (attained curriculum).;Three Illinois school districts provided enrollment records of 3,019 students in high school mathematics. Item response data from the Illinois State Board of Education (ISBE) were merged with the school district data. In order to create common curricular groups (intended curriculum) across the three school districts, a mathematics course classification system from the Council of Chief State School Officers (CCSSO) was applied based on student enrollment patterns. Thus, four curricular groups were created (General Mathematics, Algebra, Geometry, and Advanced). Opportunity to learn measures (implemented curriculum) were obtained from a group interview with high school mathematics department heads.;SIBTEST (Shealy & Stout, 1993) software was applied to the item response data to evaluate test items for DIF and item bundles for DBF. DIMTEST and DETECT software were applied to evaluate the dimensionality of the test and subtests because multidimensional items and item bundles have the potential to exhibit DIF and DBF, respectively.;Where the students are grouped based on curricular factors, more than two thirds of the test items exhibited DIF in at least one of the analyses (47 out of 70 items from 350 analyses) and about one half of the subtest analyses exhibited DBF. When the student groups had discordant coverage of item content, the magnitude of the DIF/DBF was greater. Clearly, the item, subtest, and test dimensionality differs across the curricular groups.;Psychometric analyses show that intended and implemented curriculum matters, and that it affects the student achievement (attained curriculum). Curricular implications show the need for curriculum policy makers to reduce tracking practices and to provide an enriched, rigorous curriculum to all students. |
