Do the Academic and Quantitative Literacy tests of the National Benchmark Tests have discriminant validity?

Abstract

The massification of higher education since the last decade of the 20th century has resulted in an unprecedented influx of students to universities throughout the world. In South Africa, the advent of democracy in 1994 has added impetus to this phenomenon. As a result of the poor quality of education offered at the schools they attended, however, most of the students entering universities in the country are underprepared to handle the demands of higher education in English, the language of teaching and learning at these institutions and a second language to most students. This is the case even for students who obtain good results in their high  school exit examinations. Higher Education South Africa (HESA) has  introduced a set of tests known as the National Benchmark Tests (NBTs) to assess the level of academic preparedness of the students entering universities for the first time. The NBTs comprise three tests of Academic Literacy (AL), Quantitative Literacy (QL) and Maths Literacy (ML). As their names imply, the three tests are aimed at measuring three different knowledge/skills domains that are key to student success at university. It is important therefore that performance in these tests exhibits evidence of the difference that these domains entail. The aim of this study was to  establish if two of these tests, namely, the AL and QL tests possessed discriminant validity. The results revealed that the tests lack discriminant validity.

Keywords: discriminant validity, academic literacy, quantitative literacy, test.