APPLICATION OF TEM RESPONSE THEORY (IRT) MODEL IN ASSESSING PROGRAMMING PROFICIENCY

Abstract

This study outlines a methodology for applying Item Response Theory (IRT) to objectively evaluate the programming proficiency and latent abilities of higher education students. To overcome the limitations of Classical Test Theory (CTT), item difficulty (b) and discrimination (a) parameters were calibrated using the Two-Parameter Logistic (2PL) IRT model. Experimental findings demonstrate that an IRT-based adaptive diagnostic framework significantly improves the accuracy of assessing students' actual domain knowledge compared to traditional scoring methods.