Psychometric Quality Of An Elementary School Mathematics Test: A Classical Test Theory Analysis Of Item Difficulty, Discrimination, And Distractor Functioning

Abstract

This study evaluated the psychometric functioning of a 25-item multiple-choice mathematics test on perimeter and area administered to 35 fifth-grade students. Although item analysis is widely used, evidence from routine elementary classroom assessments remains limited, particularly studies that integrate difficulty, discrimination, and option-level response patterns into actionable item-revision decisions. Using a descriptive Classical Test Theory design, all 35 students represented in the archived item summaries were included. Item difficulty (P) was calculated from the proportion of correct responses, discrimination (D) from the difference between the upper and lower approximately 27% groups (n = 10 per group), and distractor functioning from option-selection frequencies. Twenty-three items were easy (P = 0.71-0.91), while only two were moderately difficult (P = 0.69). Discrimination ranged from -0.10 to 0.60; six items showed good or very good discrimination, three had D = 0.00, and two had negative values. Distractor functioning was uneven, especially for items with concentrated response patterns. The findings indicate that the test was suitable for checking basic mastery but insufficiently balanced for differentiating students across a broader ability range. The study contributes a decision-oriented framework for retaining, revising, or reviewing classroom test items. Because the archived dataset did not contain the complete person-by-item matrix, internal-consistency reliability could not be estimated. Larger multi-school samples, repeated administrations, and Item Response Theory analyses are recommended.