Main Article Content
Die prestasievooruitsigte in verskillende universitêre vakrigtings: 'n internasionaal vergelykende ondersoek
Abstract
Verskeie statisties beduidende korrelasies wat in grootte wissel van matig tot hoog is gevind tussen 'n indeks van die prestasievooruitsigte in 14 vakrigtings aan 'n Suid-Afrikaanse universiteit en die ooreenstemmende indekse wat dekades vroeër aan twee universiteite in die VSA verkry is. Hierdie indeks het matig tot hoë, statisties beduidende, negatiewe korrelasies getoon met elk van die volgende eienskappe van hierdie vakrigtings: die gemiddelde matriekprestasies van studente daarin, die aanwesigheid van voorvereistes, en die getal geakkrediteerde vaktydskrifte daarin. Die verkreë bevindings ondersteun die algemeenheid van die verskynsel dat vakrigtings ten opsigte van prestasievooruitsigte van mekaar verskil en suggereer dat hierdie verskynsel nie uitsluitlik aan die nasienoptrede van dosente toegeskryf kan word nie, maar dat die aard van die vakrigtings ook daarmee verband hou.
Summary
Previous research unequivocally suggests that even the same group of students tends to obtain different marks in different academic disciplines. Authors such as Goldman and Widawski (1976), Elliott and Strenta (1988) and Young (1993) attribute such between-course mark noncomparability to differential grading standards, by which they mean that if the same group of students obtains a higher mean in one course than in another, the marking standards are more lenient in the former than in the latter course. In this article it is proposed that the present phenomenon could more appropriately be referred to as the varying achievement prospects in these disciplines, because the differences observed cannot be attributed solely to the marking actions of lecturers. To determine an index of the leniency or strictness of the grading standards of an academic department, Goldman and Widawski (1976) identified the groups of students that took courses in both that department and any other department, and computed for each such group the difference between their means in the two departments. They used the mean of these pairwise differences (each difference being obtained between the department in question and another department) as the index for that department. Elliott and Strenta (1988) extended this procedure to obtain such adjustment indices for grading standards between courses within departments by determining the mean of the students in a course minus the mean of the means that these students obtained in their respective sets of remaining courses in the same department. Several studies demonstrated increases in the correlations between predictors (eg, high school performance and aptitude-test scores) and university performance after the latter has been adjusted in terms of the present indices. In this study, achievement prospect indices were derived on the first-year marks of three intakes of students at a South African university. This was done by applying the Elliott-Strenta approach for 14 disciplines for whom Goldman and Widawski (1976) and Elliott and Strenta (1988) developed indices at different universities in the United States of America several decades ago. Furthermore, reasons for the present phenomenon proposed by Goldman and Hewitt (1975) were investigated by correlating objective indices of these potentially explanatory variables with the locally determined achievement prospect indices. Their suggestion that more objective facts had accumulated in the natural sciences than in the social sciences and the humanities, was investigated by correlating these indices with the number of journals in these disciplines that were accredited by the National Department of Education. Their notion that the contents of the natural sciences were hierarchically structured and hence more complex than those of the social sciences and the humanities was investigated by correlating the present indices with the presence/absence of prerequisites for the disciplines involved. The idea that students gravitated towards the disciplines that were best suited to their abilities was examined by correlating the present indices with the mean matriculation symbol point totals of students in these disciplines. Correlations of medium to large sizes (Cohen 1992) were obtained. Moreover, despite the small samples (of disciplines) involved, the majority of the results for two of the three intakes of first-year students were statistically significant, some even at the one percent level. These results support the comparability of the indices of achievement prospects obtained locally and those determined earlier in the United States and suggest that the present phenomenon cannot be explained satisfactorily in terms of the grading practices of lecturers only, but that the nature (structure, complexity, extensiveness) of the disciplines involved also needs to be taken into consideration. The present findings call for a reinterpretation and renaming of the well established finding that disciplines differ in terms of the ease with which students earn marks in them. For the present, a more neutral term, such as achievement prospects, rather than grading standards, seems to be appropriate.
South African Journal of Higher Education Vol.16(3) 2002: 139-147
Summary
Previous research unequivocally suggests that even the same group of students tends to obtain different marks in different academic disciplines. Authors such as Goldman and Widawski (1976), Elliott and Strenta (1988) and Young (1993) attribute such between-course mark noncomparability to differential grading standards, by which they mean that if the same group of students obtains a higher mean in one course than in another, the marking standards are more lenient in the former than in the latter course. In this article it is proposed that the present phenomenon could more appropriately be referred to as the varying achievement prospects in these disciplines, because the differences observed cannot be attributed solely to the marking actions of lecturers. To determine an index of the leniency or strictness of the grading standards of an academic department, Goldman and Widawski (1976) identified the groups of students that took courses in both that department and any other department, and computed for each such group the difference between their means in the two departments. They used the mean of these pairwise differences (each difference being obtained between the department in question and another department) as the index for that department. Elliott and Strenta (1988) extended this procedure to obtain such adjustment indices for grading standards between courses within departments by determining the mean of the students in a course minus the mean of the means that these students obtained in their respective sets of remaining courses in the same department. Several studies demonstrated increases in the correlations between predictors (eg, high school performance and aptitude-test scores) and university performance after the latter has been adjusted in terms of the present indices. In this study, achievement prospect indices were derived on the first-year marks of three intakes of students at a South African university. This was done by applying the Elliott-Strenta approach for 14 disciplines for whom Goldman and Widawski (1976) and Elliott and Strenta (1988) developed indices at different universities in the United States of America several decades ago. Furthermore, reasons for the present phenomenon proposed by Goldman and Hewitt (1975) were investigated by correlating objective indices of these potentially explanatory variables with the locally determined achievement prospect indices. Their suggestion that more objective facts had accumulated in the natural sciences than in the social sciences and the humanities, was investigated by correlating these indices with the number of journals in these disciplines that were accredited by the National Department of Education. Their notion that the contents of the natural sciences were hierarchically structured and hence more complex than those of the social sciences and the humanities was investigated by correlating the present indices with the presence/absence of prerequisites for the disciplines involved. The idea that students gravitated towards the disciplines that were best suited to their abilities was examined by correlating the present indices with the mean matriculation symbol point totals of students in these disciplines. Correlations of medium to large sizes (Cohen 1992) were obtained. Moreover, despite the small samples (of disciplines) involved, the majority of the results for two of the three intakes of first-year students were statistically significant, some even at the one percent level. These results support the comparability of the indices of achievement prospects obtained locally and those determined earlier in the United States and suggest that the present phenomenon cannot be explained satisfactorily in terms of the grading practices of lecturers only, but that the nature (structure, complexity, extensiveness) of the disciplines involved also needs to be taken into consideration. The present findings call for a reinterpretation and renaming of the well established finding that disciplines differ in terms of the ease with which students earn marks in them. For the present, a more neutral term, such as achievement prospects, rather than grading standards, seems to be appropriate.
South African Journal of Higher Education Vol.16(3) 2002: 139-147
