Modelling of the Advanced Level National Examination Average Pass Rate in Zimbabwe using Bayesian Hierarchical Log-logistic and Normal Mixture Approach

The national examination as one of the standard evaluation systems of education in Zimbabwe is used for the educational developments that seek to improve the quality of education in the educational sectors. This research aims to find the best model and its factors affecting the average pass rate of the Advanced Level (A-Level) national examination in Zimbabwe. Modelling was conducted using a two-level hierarchical model with factors influencing the national examination at district in the first level and those influencing the national examination provincial level in the second level. The Bayesian approaches namely hierarchical log-logistic and normal mixture were used in the modelling. The estimation of these Bayesian approaches posterior parameters was done using Markov Chain Monte Carlo (MCMC) and the Deviance Information Criterion (DIC) value was used to select the best model. The hierarchical normal mixture was found to be the best model to explain the variability of the average pass rate percentage of the A-Level national examination and all the micro and macro variables in this study significantly influenced the A-Level national examination in Zimbabwe.