<![CDATA[The quality of mathematics assessment depends not only on content coverage but also on how well each item functions empirically. This study aimed to analyze the psychometric characteristics of a fraction-based Mathematics Academic Competency Test (Tes Kemampuan Akademik/TKA) instrument for primary school students through empirical validity, reliability, item difficulty, and discrimination power. A quantitative descriptive item-analysis design was employed using responses from 27 sixth-grade students at SD Negeri Kaliboto Lor 07, Lumajang. The trial instrument consisted of 15 dichotomously scored items in multiple-choice and complex multiple-choice formats. Empirical validity was analyzed using Pearson item-total correlations with a critical value of r = 0.381 at α = .05. Reliability was estimated using KR-20, while item difficulty and discrimination were analyzed within classical test theory. Ten of the 15 items met the empirical validity criterion, whereas five were eliminated. The KR-20 coefficient was 0.70. Among the retained items, seven were moderately difficult, two difficult, and one easy; four showed good discrimination and six showed fair discrimination. The findings indicate adequate preliminary psychometric quality, although broader validation with larger, multisite samples is still required before wider use.]]>
Copyrights © 2026
