This study is to develop an improved framework for the College Scholastic Ability Test (CSAT) in Korea. To achieve this goal, the study critically reviewed the problem associated with the CSAT and the previous studies on the CSAT. This study also conducted an analysis on the predictive validity which is the validity of greatest interest for the CSAT.
The review and analysis of the previous studies on the CSAT have revealed that (1) the test characteristic of the CSAT is obscure, requiring separation of aptitude test from achievement test, (2) The CSAT's discrimination power and efficiency is not satisfactory, and that (3) the CSAT's predictive validity, namely the ability of the test to predict grade point average (GPA) at a post-secondary institution is near to zero.
Based on these reviews, the problems of the CSAT were discussed and ways of improving the present CAST were proposed.