Analysis of Multiple Choice Questions Using Classical Measurement Theory in the End of Year Assessment (PAT) of PAI and BP Class VIII Subjects at SMPN 2 Kuantan Mudik

Abstract

This study is motivated by the fact that 50% of eighth-grade students at SMPN 2 Kuantan Mudik did not meet the Minimum Competency Criteria (KKM) in the final assessment of the PAI and BP subjects. The aim of this research is to evaluate the quality and follow-up actions for the test items, analyzed using classical test theory. This research employs a mixed-method approach, combining both qualitative and quantitative methods. The sample consists of 44 students using a total sampling technique. Data collection was carried out through documentation, and data analysis used techniques applied to the final assessment (PAT) instrument. The results of this study show that, based on content validity, 94.6% of the test items met the material aspect criteria, 97.6% met the construction aspect criteria, and 100% met the language aspect criteria. An analysis of the cognitive level of the questions shows that 18 questions were categorized as C1 (Remembering), 23 questions as C2 (Understanding), 3 questions as C3 (Applying), and 6 questions as C4 (Analyzing). Furthermore, based on classical test theory, it was found that the questions had an easy difficulty level, poor discriminating power, and low distractor effectiveness. Consequently, follow-up actions recommended include revising 34 questions and discarding and replacing 16 questions