PERSON FIT ANALYSIS FOR ASSESSING ACADEMIC WRITING PERFORMANCE USING RASCH MODEL

Assessing academic writing performance is a critical aspect of educational practices. The thesis is derived from writing is an essential skill that students need to develop to succeed academically and professionally. To fulfill this, the Rasch model offers important test statistics that play a crucial role in constructing tests related to evaluation and item selection problems, as well as in decision-making regarding the generated test scores. Building upon this narrative, this research is conducted with the aim of identifying person fit in assessing academic writing performance using the Rasch model. The sample in the form of essays was obtained from a group of 40 students who had previously undergone a six-month academic writing program. The data was then analyzed using Ministep 4.8.2.0 built for Rasch Model analysis. The analysis results indicated that there were 31 students in the person misfit category, suggesting a need to reconsider the appropriateness of the conducted treatment and whether there were other factors contributing to it. Additionally, a discrepancy was found in 5 students with persons misfit or overfit, and their better performance was observed on more difficult items, contrary to the Rasch model's profile