Using the Rasch model to assess examination Beyond students´ scores

Similar to most institutions of higher learning (HLI) around the globe, the admission requirements for all HLI in Malaysia offering engineering courses include good grades in physics, mathematics or other numerate subjects. Students´ evaluations are based on the grades scored from administrative tests, quizzes and assignments are convenient as it provides a summative measure of students´ cognitive ability. However, such measure is said be lacking in precision and low in reliability when measuring achievement levels of individual students. The aim of this paper is to reassess the method of evaluation by taking into account of the measurements of items and linear measures for persons and items. Using the stochastic Rasch model, performance of the matriculation students in the Physics examination paper is examined. The study suggests that the item model gives a good fit but persons and items measurement gives a fair fit. The measurement also indicates that a number of students encounter significant difficulties in grasping some of the fundamental topics in physics. These findings provide valuable insights into the structure and functionality of the examination paper. In general terms, it provides useful information in ensuring quality delivery of teaching and learning at the matriculation level.