CEC Theses and Dissertations

Title

Predicting Success in Computer Science at a Community College

Date of Award

1997

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Graduate School of Computer and Information Sciences

Advisor

Steven R. Terrell

Committee Member

Marian Gunsher Sackrowitz

Committee Member

Susan Fife Dorchak

Abstract

This study investigated the existence of factors which might be used to predict student success in computer science at a community college. Factors examined included students' age, ethnicity, gender, and mathematics skills on entry to the college. Mathematics skills were measured by analyzing students' scores on the New Jersey College Basic Skills Placement Test (NJCBSPT). Success was determined by evaluating students' GPA in computer science courses, which was computed using students' average grade, highest grade, and lowest grade in repeated courses, and their grade in the first programming course taken. The 306 students comprising the sample group for this study were selected from full-time, computer science majors at Middlesex County College (MCC) in Edison, New Jersey in attendance from 1990- 1995. All students selected completed the NJCBSPT and took their introductory programming course at MCC. Students' age in the sample group ranged from 20 to 50 years. The average age of male students was 26.53 and the average age of female students was 27. 18. There were 200 males and 106 females. Neither age nor gender was found to be related to success in computer science. Non-Asian women were found to withdraw from computer science courses at a higher rate than male students. Five ethnic groups were represented in the sample group including 122 Asians, 115 whites, 31 Hispanics, 14 African Americans, and 2 Native Americans. An ANOVA test indicated that it is very unlikely that students of varying ethnicity have equal success in computer science. Minorities, other than Asians, were found to be less successful in computer science than whites. Using a stepwise regression, either computation or algebra skills entered into the equation first, yielding Multiple R's of not less than .30 at a p < .05 for each of the computed GPAs in the major and grade in the first programming course. A cross tabulation table demonstrated a clear trend between computation and grades in computer science and algebra and grades in computer science.

This document is currently not available here.

  Link to NovaCat

Share

COinS