Abstract
BACKGROUND: Resilience is conceptualized as an individual's ability to endure and bounce back from adversity. It serves as a crucial protective element for personal development within society, enabling individuals to navigate challenges successfully. The Brief Resilience Scale (BRS) is a widely used and freely available tool for measuring resilience. However, there is a lack of literature on its psychometric properties in the context of Indian college students. METHODS: A total of 211 Indian college students (M = 19.29 years, SD = 1.46 years) were recruited through non-probabilistic sampling to complete the Brief Resilience Scale, the Connor-Davidson Resilience Scale-25 (CD-RISC), and the General Self-Efficacy Scale (GSE). Reliability analyses of the scales were performed using Cronbach's α and McDonald's ω. Confirmatory factor analysis was used to assess model fit and test for measurement invariance. Concurrent and convergent validities were examined through Pearson's r correlations and linear and non-linear latent factor correlations. Finally, a linear regression analysis was conducted to evaluate the predictive power of the BRS and CD-RISC on GSE scores. RESULTS: The Brief Resilience Scale exhibited a unidimensional structure with satisfactory measurement invariance. It also showed satisfactory α (0.739) and ω (0.753) values. However, during validity testing, BRS's relationship with GSE (Linear: -0.119, p > 0.05, Non-Linear: RESET: 0.019, p > 0.05) and CD-RISC (Linear: -0.005-0.083, p > 0.05; Non-Linear: RESET: 0.287-0.792, p > 0.05) was noted to be insignificant when tested through linear and non-linear latent factor correlations. In contrast, these relationships were significant when tested using Pearson correlations (GSE: 0.164, p < 0.05; CD-RISC: 0.155, p < 0.05). Furthermore, the linear regression analysis revealed that CD-RISC was a better predictor of GSE than BRS. CONCLUSIONS: The results suggest that the BRS factor structure and its measurement invariance align with previous literature. However, the current sample showed no association with theoretically linked variables. As a result, its use should be limited to contexts where rigorous psychometric testing has been conducted to ensure its suitability for Indian college students. Furthermore, most prior studies have relied on Pearson correlations to assess associations. This study highlights the importance of incorporating linear and non-linear latent factor correlations to understand such relationships comprehensively.