Psychometric Properties of the Recovery Assessment Scale (RAS) in Spain: Confirmatory Factor Analysis and Metric Superiority of the 18-Item Version (RAS-18)

Abstract

Background: The adoption of the recovery model in mental health creates a need for validated instruments that go beyond the assessment of clinical symptoms. The Recovery Assessment Scale (RAS) is the most widely used test internationally. However, the need to balance the validity of psychological measurement with reduced administration time has led to the search for shorter versions. In this vein, there is an 18-item version that has not yet been validated in Spain. To study the psychometric properties of the RAS in Spain, analysing the 41-, 24- and 18-item versions, to determine the version with the best fit and greatest parsimony.

Methods: The questionnaires were administered to a sample of 307 patients with severe mental disorders (SMDs) treated in public mental health services, and a confirmatory factor analysis (CFA) was performed testing the three versions. Concurrent and divergent validity were examined using the Individual Recovery Outcomes Counter (I.ROC) scale and the Health of the Nation Outcome Scales (HoNOS) scale, respectively. A retest was also performed after 15 days to check temporal stability. 

Results: The RAS-18 demonstrated superior metrics and greater parsimony (with improvements in chi-square, Comparative Fit Index (CFI), Root Mean Square Error of Approximation (RMSEA), Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC)), whilst maintaining good results in internal consistency (alpha = 0.892; ω = 0.901), test-retest reliability (intraclass correlation coefficient (ICC) = 0.864) and convergent validity with the I.ROC (r = 0.784).

Conclusions: The study recommends the use of the RAS-18 in Spain in clinical and research contexts. This version improves the assessment of personal recovery, reducing administration time without sacrificing the psychometric accuracy of the construct.