Common item-writing flaws and psychometric properties of multiple-choice test in technical schools in Ghana

Abstract

High-quality classroom assessment is essential for effective teaching and learning, particularly in technical and vocational education, where assessment outcomes guide instructional decisions and skills development. This study investigated the common item-writing flaws and psychometric quality of multiple-choice tests (MCTs) constructed by Technical Drawing teachers in public technical schools in the Central Region of Ghana. A descriptive cross-sectional survey design was employed, involving all twelve Technical Drawing teachers using a census approach. Data were obtained through document analysis of teacher-developed MCTs and students’ response data for psychometric evaluation. Item difficulty indices, discrimination indices, and distractor efficiency were analysed using SPSS and JMetrik software. The findings revealed recurring item-writing flaws, including identifiable answer-key patterns, ambiguous wording, unclear directions, and fairness concerns related to contextual references and linguistic complexity. Psychometric analysis showed an imbalance in item difficulty levels and a high proportion of items with poor or negative discrimination indices, indicating limited effectiveness in differentiating between high-and low-performing students. Although a substantial proportion of distractors were functional, several were either non-functional or misleading, thereby weakening the diagnostic value of the items. The study highlights gaps in teachers’ application of test-construction principles and underscores the need to strengthen assessment literacy. Targeted professional development, routine item analysis, and institutional assessment quality assurance to enhance the validity, reliability, and fairness of classroom-based multiple-choice assessments in technical education.