| Mathematics placement and mathematics achievement in the community college were studied in an effort to identify a placement scheme utilizing multiple factors which would be more effective than the use of a single mathematics placement score. The study included 793 students in a Midwestern community college who placed into five different classes from Arithmetic and Prealgebra up through Beginning Algebra, Intermediate Algebra, and College Algebra.; Logistic regression was used to determine what combination of independent variables exhibited the strongest relation with mathematics achievement as evidenced by the student's final course grade. Independent variables included: COMPASS Mathematics Placement Test Score, COMPASS Reading Placement Test score, English Placement Test score, HSGPA (high school grade point average), ACTE (American College Test English score), ACTM (American College Test Mathematics score), ACTC (American College Test Composite score), age in years, and gender. Mathematics achievement was represented by final course grade, either successful (A, B, C, or “Pass”) or unsuccessful (D, F, “Progress, Re-enroll”, or “Fail”).; It was found that none of the models using backward or forward inclusion methods was a good fit to the data in any of the five courses. Indicators of goodness of fit included Model Log Likelihood (-2 LL) and χ 2, both of which were consistent and very large. This study did not confirm a predictive model of mathematics achievement using multiple measures.; This investigation involved community college students who were older and likely less homogeneous a group than the adolescents studied previously. |
