Psychometric properties of a culture-adapted Spanish version of AIDA (Assessment of Identity Development in Adolescence) in Mexico

Background: The construct "identity" was discussed to be integrated as an important criterion for diagnosing personality disorders in DSM-5. According to Kernberg, identity diffusion is one of the relevant underlying structures in terms of personality organization for developing psychopathology, especially borderline personality disorder. Therefore, it would be important to differentiate healthy from pathological development already in adolescence. With the questionnaire termed AIDA (Assessment of Identity Development in Adolescence), a reliable and valid self-rating inventory was introduced by Goth, Foelsch, Schlueter-Mueller, & Schmeck (2012) to assess pathology-related identity development in healthy and disturbed adolescents. To test the usefulness of the questionnaire in Mexico, we contributed to the development of a culture-specific Spanish translation of AIDA and tested the reliability and aspects of validity of the questionnaire in a juvenile Mexican sample. Methods: An adapted Spanish translation of AIDA was developed by an expert panel from Chile, Mexico, and Spain in cooperation with the original authors, focusing on content equivalence and comprehensibility by considering specific idioms, life circumstances, and culture-specific aspects. The psychometric properties of the Spanish version were first tested in Mexico. Participants were 265 students from a state school (N = 110) and private school (N = 155), aged between 12 and 19 years (mean 14.15 years). Of these, 44.9% were boys and 55.1% were girls. Item characteristics were analyzed by several parameters, scale reliability by Cronbach's Alpha, and systematic effects of gender, age, and socioeconomics by an analysis of variance (ANOVA). We evaluated aspects of criterion validity in a juvenile justice system sample (N = 41) of adolescent boys in conflict with the law who displayed various types of behavioral problems by comparing the AIDA scores of a subgroup with signs for borderline pathology (N = 14) with the scores obtained in the student sample using T-tests. Results: The psychometric properties of the Spanish version of AIDA proved satisfactory in the Mexican sample for items as well as scales. The reliability coefficients were alpha = .94 for the total scale "Identity Diffusion", alpha = .85 and .92 for the two primary scales "Discontinuity" and "Incoherence", and between alpha = .70 and .83 for the subscales. However, some items of the item pool in the Spanish version of AIDA did not meet all criteria for test equivalence and should thus be reformulated, taking the Mexican culture into account. Significant effects for gender and age were found. In line with our theory, the AIDA scores in the domains "Discontinuity" (high effect size) and "Incoherence" (medium effect size) were markedly higher in the delinquent boys than in the student group. Conclusion: The Spanish version of AIDA can be used in Mexico with satisfying psychometric properties, with only minor adaptions required. Our study contributes to the intercultural applicability of the AIDA instrument using the construct "identity integration vs. diffusion" as it was defined in the AIDA model for diagnostic purposes. Cultural differences, even those present in the various Spanish-speaking countries, should be modeled carefully.