Assessing Selected Predictors of Performance in Developmental Mathematics

Open Access
- Author:
- Shalyefu, Rakel-Kavena
- Graduate Program:
- Instructional Systems
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- October 27, 2004
- Committee Members:
- Barbara L Grabowski, Committee Chair/Co-Chair
John David Popp, Committee Chair/Co-Chair
Edgar Paul Yoder, Committee Member
Pui Wa Lei, Committee Member
Francis M Dwyer Jr., Committee Member - Keywords:
- developmental mathematics
mathematics performance
remedial mathematics
community college
success in math
dropouts in math
predictors of math performance - Abstract:
- ABSTRACT The purpose of this study was to investigate selected predictors (i.e. participants’ background, math prior knowledge, motivation, learning strategies, teaching strategies and practices, and instructional materials) on math performance (final grade, successful, not successful and dropouts) in an elementary algebra course of a developmental mathematics program at a community college. Two data sets were collected from two semesters (Fall 2003 and Spring 2004). Four main considerations adopted from a tetrahedral model provided a framework for analyzing the data. These included (1) learner variables (2) teacher variables (3) material variables and (4) criterial measures. Results showed positive and statistically significant correlations between intrinsic goal orientation, self efficacy for learning and performance, effort management strategies and math final grade and a negative relationship with teaching strategies and practices. The logistic regression results reproduced the correlation results with statistically significant likelihood ratios of success, not successful, and dropouts associated with class level, pre-algebraic skills, self-efficacy for learning and performance, effort management strategies and the general teaching strategies and practices in the final model. The final Logistic Regression model explained successful, not successful and dropouts with an average accuracy of 63%. The model would likely be improved with additional information about the teachers, because some of them had a statistically significant association with successful students. However, the statistically significant associations found between math performance and teaching strategies and practices were negative.