Abstract:

In Kwan et. al. (2008), the authors have proposed statistical methods to calculate the average class mark of a post-secondary subject - Quantitative Methods. In this paper, it reveals that there was a difference in the results between Quantitative Methods (QM) and Introduction to Information Technology (IIT). This difference could be explained by different student groups with different background of knowledge and experience. We know that there is a sensitivity effect: even a very small proportion of students get fail or extremely outstanding in the subject the average class mark will be influenced to a lower or higher value. By comparing the results in QM and IIT, it proves that using our proposed advanced statistical methods to find the average class marks, the sensitivity effect of extreme marks could be reasonably reduced.

References:

Wilson C.K. Kwan, Chan C. M. and Anthony W.K. Loh (2008) Application of Advanced Statistical Methods To Find The Average Class Marks
Conference proceeding of International Association for Technology, Education and Development, virtual session.

Keywords:

Average Class Mark, Scale Mixtures of Normal (SMN) Distribution, Exponential-power (EP) Distribution, Student-t Distribution, Markov chain Monte Carlo (MCMC), Gibbs Sampler, Mixing Parameter.