<![CDATA[Limited empirical evidence exists regarding the validity of instruments used to assess evaluative competence in academic libraries accreditation. The 2025 revision of the Indonesian higher-education library accreditation instrument assigns the highest weighting to service components, requiring assessors to interpret relationships across indicators. This study examined the psychometric properties of a cognitive test instrument designed to assess such competence. A total of 32 items were administered to 30 assessors, and item characteristics were analyzed using classical test theory. Results indicate highly homogeneous response patterns: 90% of respondents scored in the high category, 81.25% of items were classified as easy, 46.88% had low discrimination indices, and 65.63% contained non-functional distractors. Although the KR-20 reliability coefficient was high (0.84), internal consistency was driven by response homogeneity rather than true variance in ability. These findings suggest that high internal consistency in cognitive assessment instruments may arise from response uniformity rather than reflecting actual variation in underlying ability, particularly in professional evaluation contexts.]]>
Copyrights © 2026
