Die verband tussen matriekprestasie en eerstejaarprestasie vir opeenvolgende innames aan dieselfde universiteit
AbstractDie korrelasie tussen matrieksimboolpunttotaal (MST) en die gemiddelde kurrikulumpersentasiepunt (GKMP) is vir verskillende opeenvolgende eerstejaar-innames van wit mans, wit vroue, swart mans en swart vroue aan dieselfde universiteit in die negentigerjare bereken, sowel vóór enige aanpassings in GKMP, as ná aanpassings vir verskille in puntoekenningstandaarde en kursuslading. Die onderhawige korrelasies was deurgaans hoër vir wit as vir swart studente en in die meeste gevalle hoër vir mans as vir vroue. Soos op grond van 'n makliker bereikbare MST-afkappunt en die opheffing van die semesterpuntvereiste as toelating tot eksamens verwag kon word, was daar 'n toename in die onderhawige korrelasies vir wit studente. Die korrelasies vir swart studente het die meeste baat gevind by aanpassings vir kursuslading terwyl die korrelasies vir wit studente die meeste voordeel getrek het uit aanpassings vir punttoekenningstandaarde.
Summary Admissions to South African degree programmes other than the professionally oriented programmes such as medicine, social work, and nursing, appear to present no problem to matriculants with university exemption. However, this situation is likely to change if the White Paper proposal of the Department of Education (1997) to subsidise only a limited number of vacancies for particular degree programmes at particular institutions, is to be implemented. Typically, admissions research entails investigations into the correlation between one or more predictors (eg, the matriculation symbol point total (MST)) and a criterion of university performance (eg, the mean percentage mark (MCPM) over the courses taken) for a particular sample (intake) of students at a particular institution of higher education. Although interest is usually centered on the predictor(s), the obtained results also may depend on the criterion used and the sample involved (more particularly, its selectiveness). Universities are encouraged to regularly undertake validity studies to determine whether any particular trend in the above predictor-criterion correlation is discernible over a period of time. The 1990s present a particularly significant transition period in South African higher education in general, and at the University of the Orange Free State (UOFS) in particular. During this period several changes took place in the composition of the intake of students, the comparability of MST as predictor for different demographic groups and the strictness of admission to university examinations and hence the heterogeneity of university performance at the institution concerned. Probably the most striking feature at this University as far as enrolments were concerned, was the increased number of registrants from educationally disadvantaged high schools. As is true of all correlations, the present correlation depends on the heterogeneity of the sample on which it is based. Findings reported by Van Rooyen and Huysamen (2000) suggest that during the 1990s it became easier to obtain an MST with exemption so that the present institution has become less selective even though it did not lower the MST cutoff for admission. Secondly, as students from both historically white and historically black high schools in any given province have been writing the same matriculation examination since 1997, the MSTs of black and white applicants from a given province should have become more comparable than before the institution of common examinations - a change which was expected to benefit the present correlation. As far as the criterion measure was concerned, a semester mark of 50% or higher was no longer required for admissions to examinations at the UOFS as from 1997. The relaxing of this regulation may be expected to have increased the heterogeneity of the MCPM of subsequent intakes, and hence to have affected to MST-MCPM correlation positively. By contrast, the differing course loads of black students in particular are likely to have lowered this correlation. The present study was aimed at examining the effects, if any, of these changes on the present correlation from 1992 to 1999 for white and black students. It set out to correlate MST with each of the following: MCPM, the MCPM weighed by the credits carried by the courses taken, the MCPM adjusted for marking standards, and the MCPM weighed by the credits carried by the courses taken after the latter had been adjusted for marking standards. These correlations were calculated separately for white men, white women, black men, black women, white students and black students for each of the intakes of 1992, 1994, 1995, 1996, 1997, 1998 and 1999 at the UOFS. The results suggested a slow increase in these correlations until 1995 for white men and until 1996 for white women, a decrease for white women afterwards and some fluctuations for white men. For black students a small increase after 1998 was observed although it was considered too early to conclude whether this was a permanent change. In agreement with findings in the United States of America the correlations were higher for white than for black students, but contrary to American findings, they were in most cases higher for men than for women. Whereas the correlations for white students benefitted most from corrections for differences in marking standards, those for black students improved most after they had been corrected for differing course loads. The results demonstrated the advantages of (I) performing validation studies such as the above on a regular basis, (ii) computing the present correlation for different performing subgroups separately, and (iii) performing corrections for differing marking standards and for differing course loads.
South African Journal of Higher Education Vol.15(3) 2001: 142-149