Academic performance of English first and second language students: selection criteria

The effects and consequences of using different kinds and combinations of s tudent selection criteria are reported. The matriculation points score, a t est of language and mathematics competence, a score based on tutorial assig nments, matriculation mathematics, and a combined criterion, constituted th e selection criteria. The first semester examinations of a group of first-y ear psychology students (n = 543), consisting Of English first (n = 340) an d second (n = 203) language speakers, constituted the outcome against which the selection criteria were compared:The five different selection criteria were applied retrospectively, thereby yielding comparative data on the eff ects of each specific criterion. Using this procedure, it-was possible to c ompare false-positive errors:(validity), false-negative errors (equity), to tal error values (accuracy), exclusion rates (access) and failure rates (ex cellence), for each of the selection criteria. The results show that the cr iteria yielded a wide range of different effects and that the effects were not uniform for English first- and second-language students. In particular, English second-language students were disadvantaged by the use of the matr iculation points score, especially with respect to access and equity. Altho ugh relative to the matriculation points score, the test of language and ma thematics competence yielded increases in access and equity for English fir st-language students,it yielded the opposite effects for second-language st udents for whom a lower failure rate was offset by the highest rate of excl usion. The results indicate that the choice of a selection criterion will h ave consequences, intended or not, that reflect admission policies, whether explicitly or implicitly acknowledged.