Using Standardized Tests to Predict Achievement in an Introductory High School Computer Course

Abstract

A number of studies conducted with college level students have found that mathematics abililty is a key predictor of success in computer science courses. In addition, several studies have indicated that standardized computer programming aptitude tests have low predictive validity. In order to investigate how these findings apply to high school students, 61 students in an introductory computer science course were administered the Iowa Test of Educational Development (ITED), Science Research Associates (SRA) Inc.'s Computer Programming Aptitude Battery and the Wolfe Programming Aptitude Test. First quarter, second quarter and semester test point totals were used as dependent variables. It was found that ITED's Quantitative Thinking (QT) scale best fit the dependent variables. Furthermore, an optimal QT mastery score was established, tested and found to be statistically valid. The results, like those of previous studies, indicated that mathematics ability is an important factor in predicting success in computer science.