Date of Award

Fall 12-2016

Degree Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Center for Science and Math Education

School

Center for Science and Math Education

Committee Chair

Richard S. Mohn

Committee Chair Department

Educational Research and Administration

Committee Member 2

Sherry S. Herron

Committee Member 2 Department

Center for Science and Math Education

Committee Member 3

Kyna Shelley

Committee Member 3 Department

Educational Research and Administration

Committee Member 4

Deborah B. Booth

Committee Member 4 Department

Chemistry and Biochemistry

Committee Member 5

Sungwook Lee

Committee Member 5 Department

Mathematics

Abstract

ACT scores are widely used to predict outcomes in coursework and serve as placement guidelines for college level courses such as college algebra. Due to a changing college environment, the appropriateness of these placement decisions takes on a new, more critical light. Given the rate of success for current predictions in college algebra, and the resulting consequences for misplacement, this study examines the predictive potential of the ACT scores through the examination of non-linear variables and a metaphorically chaotic interpretational lens.

The literature base for this study reveals, almost to exclusion, the use of linear models for the prediction of success in college algebra. This tendency dates back to the late 1920’s. While scattered references, and a single doctoral study, have suggested the use of non-linear variables as a viable prediction method, the topic has seen little emphasis in the last 50 years. Using this as a basis for examination, and a metaphorically chaotic interpretational lens based on the non-linearity of social constructs, this study focuses on the use on non-linear combinations of the ACT sub-scores as variables in regression models to predict the outcome of college algebra classes conducted over a two year period at Jones County Junior College.

Utilizing the techniques of enumerative combinatorics, this study focuses on a set of 69 variables developed through non-linear combinations of the ACT sub-scores. An additional set of general college readiness variables were also developed as part of the metaphorically chaotic interpretational lens. These variables were subjected to a series of statistical analyses to determine the most suitable non-linear variables for inclusion in the models. Serving to provide both focused and broad examinations of college algebra outcome predictions, these models were compared to the base models currently in use at academic institutions in the State of Mississippi.

ORCID ID

0000-0001-5939-8859

Download

Share

COinS