Rasch analysis of the 10-item Connor-Davidson Resilience Scale among Chinese patients with major depressive disorder

Compared to Classical Test Theory, Rasch analysis could provide more precise and specific estimates for the quality of measures. This study aimed to conduct Rasch analysis to examine the psychometric properties of the 10-item Connor-Davidson Resilience Scale (CD-RISC-10) in Chinese patients with major depressive disorder (MDD). The sample was comprised of 136 participants (66.9 % females). Severity of depression was assessed through the Hamilton Depression Rating Scale 17-item and resilience through the CD-RISC-10. Rasch analysis was employed to evaluate unidimensionality, local independence, item fit, reliability, differential item functioning (DIF), ordering of response category and targeting. Rasch analysis confirmed the unidimensionality of the CD-RISC-10. All items demonstrated adequate fit, and all items were locally independent. No evidence of disordered category or threshold of rating scale was observed, indicating reasonable response category setting. There was no significant DIF except for item 1 and item 6 concerning sex. Moreover, all items were well-targeted and patients with moderate levels of resilience were better targeted. The CD-RISC-10 demonstrated acceptable psychometric properties among Chinese patients with MDD, providing more precise estimations for patients with moderate levels of resilience than those with high and low levels.