An investigation into the predictive validity of the CAP test.

Abstract

The Vocational Training landscape in South Africa changed in 2000 when the 153 Technical Colleges merged to form 50 macro Further Education and Training (FET) Colleges with new vocational programmes. In an endeavour to assist prospective students in making an appropriate choice of programme, placement tests prior to registration were implemented. The Competency and Placement (CAP) test was developed to identify gaps in the students’ knowledge of the grade 9 curriculum in order to identify high risk students prior to registration. This study aims to discover whether results of the CAP test of students at Boland Further Education and Training (FET) College correlate with their final year-end results in numeracy and literacy, and to what extent the test could be used to predict students’ future achievement. Data from 710 students were used to determine the correlations between their CAP scores and their year-end examination results. A Pearson product-moment coefficient (r) was calculated to indicate the extent of correlation for different groups. The coefficients of determination (r²), a coefficient used in the prediction of future outcomes, were also calculated. It was found that there were correlations of different strengths for different groups, but the coefficients of determination did not indicate that the CAP test could be used to predict the outcomes of the final year-end examinations. Therefore it can be concluded that the CAP test is not a valid instrument to predict academic success and should be used only to identify grade 9 mathematics and English skills gaps, according to the purpose of its design.