Nie-kognitiewe voorspellers van akademiese prestasie by eerstejaar-universiteitstudente

English: From several quarters appeals are made for greater tertiary-educational access for previously disadvantaged commumtres. This has an enormous impact on universities. South Africa possesses limited financial resources and cannot spend large sums of money on students with little or no chance of academic success. Selection is thus necessary for the sake of the university and the student. Traditionally, aptitude tests and high school performance have been used fairly successfully in the identification of students with a good chance of achieving success at a tertiary-educational level (Huysamen & Raubenheimer, 1999). The fairness of these selection processes is strongly questioned in view of black learners' inadequate schooling (Miller, 1992). It is specifically in this area that the use of non-cognitive variables may be of value. Tracey and Sedlacek (1984, 1985, 1987) found in various studies that non-cognitive variables as measured by their Non-Cognitive Questionnaire (NCQ) may play an important role in the prediction of academic achievement. Locally, however, there appears to be a shortage of research to evaluate the value of non-cognitive variables as part of an admissions test battery. This study aims to determine firstly whether the accuracy of predictions concerning black students' academic achievement may be increased by utilizing both cognitive and non-cognitive variables. Secondly it investigates whether noncognitive variables make a greater contribution to predicting the tertiary academic success of black students as compared to their white counterparts. The different selection methods used at South African universities, as well as the problems experienced with regard to applicants from diverse usefulness of non cognitive factors. Moreover, the seven non-cognitive dimensions upon which the NCO is based are discussed in greater detail. The value of study methods, attitude and motivation are also discussed. Two samples were used in this study. The first group consisted of 475 first-year students who completed the NCQ. The second sample consisted only of black first-year psychology students. The NCO as well as the Study Methods Questionnaire (Entwistie et al., 1971) were administered to the second sample. The first-year students' academic achievement during the first semester was used as the dependent variable. The following predictors were used in this study: matriculation marks, NCO and Studymethods. A hiarchical (multiple) regression analysis was performed in order to determine whether the above-mentioned noncognitive variables accounted for any criterion variance beyond that predicted by matriculation marks. The results of this study showed that the cognitive and non-cognitive variables for the white students collectively explained 28.69% of the criterion variance. The cognitive variable accounted for 27.24% for the variance whereas the noncognitive variables accounted for 8.4% additional criterion variance bringing the percentage of explained variance of the full model to 8.98. The latter is significant at the 1% level. The results suggest that the non-cognitive variables account for a greater portion of the variance in tertiary academic achievement among black students than does the cogntive variable of matriculation achievement. The former is also proportionally higher among black than among white students. In the second sample it was found that non-cognitive variables accounted for 12.95% of the criterion variance in addition to the 0.89% explained by matriculation marks. Individually none of the non-cognitive variables could account for any variance at a statistically significant level.