Attempt to validate the Self-Construal Scale in French: Systematic approach and model limitation

Introduction.-Independent and interdependent self-construals are included in individuals' self definitions. The 24-item Self-Construal Scale (SCS) was developed by Singelis (1994) as a means of measuring the "two selves" of individual identity, namely the independent self and the interdependent self. It has been translated into a number of different languages including French. Yet, proper psychometric validation procedures of the scale in foreign languages are lacking which is problematic given the recurrently reported subscales poor reliabilities. Objective.-The aim of the present paper is to present the steps followed in order to validate a French translated version of Singelis' (1994) 24-item Self-Construal Scale. Following such a systematic validation approach enables us to locate psychometric weaknesses and assess to what extent a standard validation procedure can address these limitations. Method.-Study 1 pertains to the translation of the inventory, item face-validity checks, and factor analyses. Study 2 aims to assess the inventory's test-retest stability, as well as its criterion-related validity based on correlations with Big Five personality traits. Results.-Results evidenced that back-translation, face-validity check and item selection did not enhance the SCS to a valid psychometric level. Factor analyses revealed that a three-factor model proved a better fit with the collected data. Conclusion.-Given the poor psychometric properties of the SCS and the emergence of theory refinements, future research should consider alternative conceptualizations of self-construal. (C) 2016 Elsevier Masson SAS. All rights reserved.