Validity issues in higher education selection: A Swedish example

Abstract

Applicants to higher education find certain courses and programmes more attractive than others. When the number of applicants to a course or programme exceeds the number of available places, a selection has to be made. The main selection instruments in Sweden are grades from upper secondary school, SweSAT scores (Swedish Scholastic Aptitude Test) and various types of course-specific selection instruments for programmes like architecture, journalism and medicine (H6gskoleverket, 1997a). The quality of the selection process has been evaluated mainly by thorough studies of the predictive validity of the selection instruments, that is, by examining the correlation between the admitted students' results on the chosen selection instrument and an indicator of subsequent academic performance. It is important to observe that the predictive validity of the selection instruments is a critical factor but is not in itself sufficient for an assessment of the validity of a selection. In light of the changed concept of validity (see e.g., Cronbach, 1988; Messick, 1989; Shepard, 1993; Wolming, 1998) predictive validity is regarded as only one of several critical factors considered in the assessment of the validity of a selection instrument. In contrast, other questions connected with the concept of validity have assumed greater importance: What do the instruments actually measure? Are the criteria for the assessment of student performance relevant? What are the consequences of the use of a