Maximizing Learning Outcomes: A Comparative Analysis of IRT- and CTT- Differentiated Learning based Design

Challenges in addressing diverse student abilities often hinder effective learning, particularly in complex subjects like linear equations and inequalities. This research aimed to compare the effectiveness of Item Response Theory (IRT)-based and Classical Test Theory (CTT)-based differentiated learning designs in improving student performance in linear equations and inequalities. Conducted in two secondary schools, the study involved 126 students, with 61 students in the IRT group and 65 students in the CTT group. A quasi-experimental design with pretest-posttest control groups was employed to assess learning progress. The results showed that both IRT and CTT-based learning interventions led to significant improvements in student performance. However, the IRT-based approach, which grouped students based on their individual ability levels and tailored tasks to their proficiency, resulted in a significantly higher average posttest score and a very large effect size. The CTT-based approach also showed improvement but with a smaller effect size. The findings suggest that IRT offers a more precise and effective method for differentiating instruction, leading to better learning outcomes, particularly in complex subjects like linear equations. This study underscores the potential of IRT in enhancing educational practices and improving student learning outcomes.