ANALYSIS OF CLUSTERED MEASUREMENTS: A COMPARISON OF THE PERFORMANCE OF FOUNDATION YEAR STUDENTS, 1994 COHORT WITH THOSE OF DIRECT STUDENTS, 1995 COHORT, AT THE UNIVERSITY OF LIMPOPO, SOUTH AFRICA

摘要

In applied sciences, one is often confronted with the collection of correlated data. Dependence between observations from the same study subject renders invalid the usual chi-square tests of independence and inflates the variance of parameter estimates. Disaggregated approaches such as hierarchical linear models which are able to adjust for individual level covariates are suitable in the analysis of such data. We compare the performances of Ex-UNIFY (Foundation Year) students and direct students at first year of Bachelor of Science programme at the University of Limpopo, South Africa. Both aggregated and disaggregated analyses are performed. The results show that Ex-UNIFY students perform better than direct students in their main examination of first year in their Bachelor of Science programme. Disaggregated analysis is able to adjust for individual level covariates, therefore gaining power over aggregated analysis.