The Comparability of General Scholastic Aptitude Test Scores across Different Population Groups

Abstract

The construct validity of the General Scholastic Aptitude Test (GSAT) for certain sub-populations was investigated. A one-factor structure was indicated for all groups considered. Differences between the factor structures of brown and white Afrikaans-speaking children were unacceptably large. It was shown that differences were related to level of achievement. With scholastic achievement in Physical Science and History as criteria, slight differences were observed between the regression lines of these groups. Differences could again be attributed to the general level of achievement of the groups. The findings are interpreted against the background of important differences between the experiential and ecological contexts in which the members of these sub-populations function.