SYCHOMETRIC ANALYSIS OF INFORMATICS TEST FOR TEACHERS OF TECHNICAL AND VOCATIONAL EDUCATION

Abstract

This article presents the results of the validation of an informatics test used for the knowledge assessment of Technical and Vocational Education and Training (TVET) teachers in the Republic of Kazakhstan. The legal and regulatory framework for teacher attestation, the structure of qualification categories, and the testing procedure are examined. The empirical base consisted of data from 112 teachers who completed 50 dichotomous items. Psychometric analysis was conducted using Classical Test Theory (CTT) and the Rasch model (IRT). Moderate test reliability (KR-20 = 0.655; α = 0.65) and average item difficulty (0.434) were established. Items with excessive difficulty (11 items, p ≤ 0.2), excessive ease (1 item, p = 0.89), negative discrimination (4 items), and low discrimination (6 items) were identified. IRT analysis confirmed the misfit of several items to the model: negative correlation with the overall ability measure (No. 1, 13, 37), deviations in fit statistics (Outfit MNSQ >1.3 for No. 10;

<0.7 for No. 35). The Wright Map revealed a lack of items with high difficulty. Based on the analysis, recommendations are proposed for revising problematic items, strengthening distractors, and conducting re-approbation. It is concluded that the test demonstrates content validity and is suitable for objectively assessing teacher competence after implementing the proposed corrections.