Adaptation and psychometric validation of the Hindi version of the self-compassion scale

Abstract Background The current study reported an adaptation of the Self-Compassion Scale (SCS) in Hindi and then tested its factorial structure, reliability, and validity to produce a 26-item scale. Methods The study was conducted on a sample of 843 university students (males = 138, females = 695, gender not disclosed = 10) aged between 18 and 26 years (mean = 19.54 years). Confirmatory Factor Analysis (CFA) and Exploratory Structural Equation Modelling (ESEM) assessed a one-factor, a two-correlated factor, a six-correlated factor, a single-bifactor, and a correlated two-bifactor model. Results Results suggest that the single bifactor and six-correlated factor models (most notably the ESEM) were the best-fitting models for representing the multidimensional structure of self-compassion. Content validity indices, along with expert ratings of the appropriateness and comprehensibility of the translated statements, supported the adequacy of the scale at the expert-review level in this cultural context, yielding items with satisfactory inter-rater reliability in this study. Results also demonstrated that the SCS had convergent validity, with moderate to strong correlations with expected constructs (i.e., satisfaction with life, subjective happiness, self-esteem, and positive affect as well-being indicators) and discriminant validity as indicated by an insignificant relation with unrelated constructs (i.e., career adaptability and social desirability), confirming the specific measurement domain of the SCS. The scale exhibited good test-retest reliability (over a four-week time frame) in the obtained sample, supporting its use for longitudinal research. Conclusions These findings indicate that the self-compassion scale Hindi version (SCS-H) is an instrument for assessing the effect of self-compassion on well-being in Hindi-speaking adults. Given the comparatively weak convergent validity of the Self-Judgment and Over-Identification subscales, scores on these two subscales should be interpreted with caution.