Predicting different grades in different ways for selective admission: disentangling the first-year grade point average

The first-year grade point average (FYGPA) is the predominant measure of student success in most studies on university admission. Previous cognitive achievements measured with high school grades or standardized tests have been found to be the strongest predictors of FYGPA. For this reason, standardized tests measuring cognitive achievement are widely used as a tool for selective admission to higher education. The FYGPA, however, measures many markedly different aspects of student success. In this article it is shown that when the FYGPA is divided into averages that reflect performance on different types of goals, the predictive value of previous cognitive achievement differs significantly between these disentangled averages. It is therefore important to distinguish between different types of goals when considering what student success is, and which students should be admitted to particular university programmes.