Methodological evaluation of the cross-cultural validation of psychiatric scales in Tunisia: about 43 doctoral theses in medicine

BACKGROUND: The cross-cultural validation of psychiatric scales is an essential step to ensure the relevance of assessment tools in specific cultural contexts. In Tunisia, several medical doctoral theses have focused on this field, but their methodological quality has not yet been systematically evaluated. This work aimed to assess the methodology of Tunisian medical theses dealing with the cross-cultural validation of psychiatric scales. METHODS: We conducted a study evaluating Tunisian theses on the cross-cultural validation of psychiatric scales. The analysis focused on translation steps as well as psychometric validation methods, using an evaluation grid designed based on international recommendations. RESULTS: A total of 43 theses concerning the validation of 50 scales were included in our analysis. For translation, direct translation was performed for all scales, while back-translation was applied to 43 scales. Indeed, during the translation process, an expert committee concerned 28 scales, whereas a pretest was administered for only 17. Regarding sampling to ensure psychometric validity, the size of the studied populations ranged from 58 to 2577 participants, and 37 scales used non-probabilistic sampling. An expert committee evaluated face validity for 22 scales and content validity for 42 scales, whereas a pretest was applied to 10 scales for face validity and 40 for content validity. On the psychometric level, internal consistency was generally satisfactory, with most alpha coefficients above 0.70. Temporal stability was examined in 29 scales, often using the ICC. Internal construct validity was studied for 38 scales, mainly through exploratory factor analysis and principal component analysis. External construct validity was less explored, with convergent validity evaluated in 23 scales and divergent validity in only 5. Finally, only 17 scales resulted in publications in indexed journals. CONCLUSIONS: This study highlighted certain gaps in the scale validation process, particularly in some translation and psychometric validation steps. Standardization of procedures strengthened psychometric training, and better scientific dissemination of results are essential to ensure the reliability and usability of instruments adapted to the Tunisian context.