Evaluation of reliability generalization of conner-davison resilience scale (CD-RISC-10 and CD-RISC-25): A meta-analysis

Abstract

Background: Resilience, a critical multi-faceted construct in psychological research, is often measured using Conner-Davison Resilience Scale (CD-RISC-10 and CD-RISC-25). This reliability generalization (RG) meta-analysis delves into evaluate the level of reliability generalization estimate of both CD-RISC-10 and CD-RISC-25 in assessing resilience across diverse populations and settings. Methods: A reliability generalization meta-analysis on the psychometric properties of CD-RISC-10 and CD-RISC-25 was conducted, encompassing 27 studies. The original versions' psychometric properties were systematically retrieved from databases including PubMed, PsycINFO, Google Scholar, Web of Science, Science Direct, and Scopus, with a focus on studies published between 2018 and 2023. The study protocol, including the specific methods for the reliability generalization meta-analysis, was pre-registered in the Prospero database (registration number CRD42023479052). This pre-registration ensures transparency and minimizes the risk of bias in the study design and analysis. Results: The analysis revealed a combined estimated overall estimate of Cronbach's Alpha of 0.89 (95% CI [0.87, 0.91], z = 77.20, p < 0.05), indicating a high level of reliability for CD-RISC-10 and CD-RISC-25. CD-RISC-10 exhibited an overall estimate of Cronbach's Alpha of 0.8732 (95% CI [0.85, 0.10], z = 69.81, p < 0.05), indicating a high level of reliability, while CDRISC- 25 also demonstrated an overall estimate of Cronbach's Alpha of 0.8922 (95% CI [0.87, 0.91], z = 77.20, p < 0.001), indicating a high level of reliability. Furthermore, CDRISC- 10 displayed commendable reliability (ωα+ = 0.86), slightly lower compared to the impressive reliability of CD-RISC-25 (ωα+ = 0.89), with a significant difference (t = 0.1159, p > 0.001). The mixed-effects model revealed a non-significant moderating effect of the CDRISC language version on reliability estimates (coefficient = -0.0017, p <0.05). Conclusion: The results affirm the high overall reliability of both CD-RISC-10 and CD-RISC-25, with CD-RISC-25 exhibiting a slightly superior level. The non-significant moderating effect of language version suggests that the psychometric properties of these scales remain robust across different linguistic adaptations. These findings enhance our understanding of the CD-RISC scales, providing practitioners, researchers, and clinicians valuable insights for informed scale selection in diverse contexts. The commendable reliability of both scales underscores their utility in assessing and promoting resilience across varied populations and settings. Future research should explore specific contexts, demographics, and applications, enhancing their utility for diverse populations and settings.