METHODS: A cross-sectional study involving 143 participants was conducted at an urban teaching hospital, an urban government hospital, and a rural government hospital. Following the translation of the RKI, its internal reliability was determined using Cronbach's alpha. Construct validity was also determined using confirmatory factor analysis.
RESULTS: The Malay-Version RKI (RKI-M) has good internal reliability with a Cronbach's alpha of 0.83. However, the Malay-version RKI failed to replicate the original four-factor structure. The final model only achieved the best model fit after the removal of 9 items with two-factor loadings: (GFI = 0.92; AGFI = 0 0.87; CFI = 0.91; RMSEA = 0.074).
CONCLUSION: The 20-item RKI-M is reliable but has poor construct validity. However, the modified 11-item Malay-version RKI is a more reliable measure as it has good construct validity, with room for future studies to examine the psychometric properties of the modified 11-item RKI among mental health care workers. More training on recovery knowledge should be done, and a simple worded questionnaire should be developed in keeping with local practitioners.