PSYCHOMETRIC VALIDATION OF A COMPUTERIZED TUNISIAN ARABIC AMT-AI IN A NON-CLINICAL SAMPLE

Abstract : This study examined the psychometric adaptation of the computerized Tunisian Arabic Autobiographical Memory Test – Alternating Instruction (AMT-AI) in non-clinical university students. Study 1, an exploratory study with 249 participants, evaluated item-level structure, memory performance, and preliminary convergent validity. Participants produced higher specific memory rates in the AMT-Specific retrieval-only condition than in the AMT-AI alternating-instruction condition, whereas categoric memory rates remained stable across conditions. Exploratory factor analysis supported a unidimensional structure, and internal consistency was good across indices. Preliminary convergent validity showed negative associations between AMT-AI specificity and depressive symptoms and maladaptive cognitive emotion regulation strategies, with modest positive associations for adaptive strategies. Study 2, a confirmatory study with 300 participants, tested the factor structure using confirmatory factor analysis and showed excellent model fit, while replicating internal consistency and convergent validity patterns. These findings indicate that the AMT-AI reliably captures autobiographical memory specificity, preserves expected retrieval patterns under alternating instructions, and is suitable for use in Tunisian university populations. Keywords : AMT-AI; Autobiographical memory; Memory specificity; Psychometric validation; Rumination.