Evaluating the Level of Difficulty of University Statistical Course Assessments: A Modern Perspective through Item Analysis

Abstract

University statistics courses serve as critical platforms for developing learners' mathematical thinking, problem-solving skills, and analytical abilities. To gauge the effectiveness of these courses and to provide learners with a fair and accurate assessment of their understanding, it is imperative to design assessments that appropriately balance the level of difficulty. Striking the right balance between challenging questions that stimulate critical thinking and accessible questions that gauge fundamental knowledge is a complex endeavor. In this context, item analysis, a data-driven technique for assessing the performance of individual test items, emerges as a valuable tool to ensure the quality and fairness of statistics assessments in higher education. In this study, we seek to explore the contemporary applications of item analysis in the evaluation of the level of difficulty in university statistics course assessments. With the rapidly evolving landscape of educational research and the advent of advanced statistical methodologies, this study aims to provide a modern perspective on item analysis techniques that can empower educators to create more effective statistics assessments. In this paper, we use the facility index and discrimination index to compare the assumed difficulty level and expected difficulty level of questions in the final exam of an advanced diploma course at the University of Technology and Applied Sciences (UTAS), Nizwa, Sultanate of Oman.