Authors : N. Lohgheswary, S. Salmaliza and H. Othman
Abstract: Assessing students through final exam questions is a common assessment method at tertiary level. The overall grade will classify the students into categories. Yet, this overall grade does not show how the student answered the exam questions. The Rasch Model can show the ability of each student answering the exam questions. It also gives information on which is the most difficult and the easiest question in an examination. This study illustrates the steps of the exam marks that are processed in the Rasch Model. This begins by entering raw exam marks in the Excel worksheet. Then, the marks will be normalized and changed to the Likert scale. Then the data is saved as formatted text. Lastly, it will run in WINSTEPS. Summary statistics for person, summary statistics for item, fit statistics, item dimensionality and person-item distribution map are among of the Rasch Model output that can be obtain from WINSTEPS. These outputs show the performance of the students and the difficulty of the exam questions. The Rasch output can also identify a misfit examination question. This means the question is extremely hard for the students to answer. This study will be helpful to a new user to use Rasch Model because it will help him or her to analyze any data, especially, examination questions.
N. Lohgheswary, S. Salmaliza and H. Othman, 2019. Transferring Raw Data for Rasch Model Analysis. Journal of Engineering and Applied Sciences, 14: 7846-7852.