STATISTICAL VALIDATION OF AN ENGLISH FOR SPECIFIC PURPOSES ACHIEVEMENT TEST USING CLASSICAL TEST THEORY: A CASE STUDY OF A HIGHER EDUCATION INSTITUTION

Abstract

Language achievement tests are widely used in higher education to assess student learning outcomes; however, their psychometric quality is rarely examined. This is particularly common in professional technical education, where teachers often design tests independently without conducting statistical validation. This article presents a Classical Test Theory (CTT)-based validation of a Professional English achievement test administered to first-year students in the Technical Maintenance of Automobiles programme at Šiaulių valstybinė kolegija, Lithuania. The study was motivated by a professional development partnership between the institution and Universidad de Jaén, Spain, within the NEOLAiA European Universities Alliance. A multiple-choice pilot test was analyzed using IBM SPSS Statistics. The analysis focused on item facility values, distractor functioning, item discrimination, and internal reliability. The results indicated that 80% of items fell within the acceptable facility value range, 45 out of 50 items demonstrated acceptable to excellent discrimination, distractors functioned effectively, and overall internal reliability was satisfactory (Cronbach’s α = .783). The findings confirm that the test demonstrates adequate psychometric quality and highlight the practical value of CTT as an accessible tool for institutional test validation.