Psychometric Properties of the Toronto Alexithymia Scale Using the Rating Scale Model Among a Sample of Thaiz University Students

Abstract

This study aimed to examine the psychometric properties of the Toronto Alexithymia Scale (TAS-20) among a sample of Taiz University students, using the Rating Scale Model. The sample consisted of 284 students from the Faculties of Arts and Engineering, selected through stratified random sampling. Results indicated that the scale satisfied the model's underlying assumptions, including unidimensionality within each subscale, local independence, and proper threshold ordering. All twenty items demonstrated adequate fit to the model, with the exception of a single borderline item that did not warrant removal. The scale further demonstrated high measurement precision around the mean trait level, with reduced precision at extreme levels, while item and person separation reliability indices were high across all three subscales. Differential item functioning analysis revealed no item exhibiting practically significant gender-based bias. The study recommends adopting the scale within the Yemeni context, scoring its three subscales separately rather than as a composite score, and exercising caution when interpreting extreme scores in individual clinical applications.