Analisis Aitem Subtes Numerik UGRT pada Tes Seleksi masuk Pendidikan Profesi Psikologi Universitas Sumatera Utara

Abstract

This study is motivated by the importance of evaluating the USU General Reasoning Test (UGRT) as a selection tool for admission to the Psychology Professional Education Program at Universitas Sumatera Utara (USU), particularly in the numeric subtest. The UGRT is used to measure prospective students' cognitive reasoning abilities; however, its psychometric evaluation has not yet been comprehensively conducted. The objective of this research is to analyze the reliability, item discrimination, and item difficulty level of the UGRT numeric subtest. This study employs a quantitative approach using Classical Test Theory (CTT) analysis. Data were obtained from the admission test results of the Psychology Professional Education Program at USU. Data analysis includes calculating the reliability coefficient, item discrimination index, and item difficulty level. The findings indicate that the reliability of the numeric subtest, as measured by Cronbach’s Alpha, is 0.639, suggesting a moderate level of reliability but not fully stable in measuring the intended construct. Most test items exhibit low to negative discrimination indices, and the majority are categorized as difficult based on the difficulty index. These findings highlight the need for item revision to enhance the validity and reliability of the UGRT as a selection tool.