Descriptive statistics of the RRI-C items.

The current study assessed the psychometric properties of the long (24 items) and brief (12 items) versions of the Real Relationship Inventory–Client (RRI-C) in a United States sample. The RRI-C is the most used quantitative measure of the real relationship construct, yet its psychometric properties have not been explored outside its development studies. A sample of 700 adults in individual psychotherapy was recruited in the study and filled out a comprehensive battery of measures. Analytical techniques included confirmatory factor analysis (CFA), exploratory structural equation modeling (ESEM), multigroup CFA, multigroup factor analysis alignment, item response theory, internal reliability assessments, Bland-Altman regression analysis, and calculation of reliable change benchmark thresholds. Both RRI-C versions demonstrated a bifactor structure encompassing Genuineness and Realism dimensions. The bifactor ESEM model provided strong fit: χ2[210] =  482.464, CFI =  0.999, TLI =  0.998, RMSEA =  0.043, SRMR =  0.020 for the 24-item RRI-C; χ2[45] =  111.916, CFI =  0.999, TLI =  0.998, RMSEA =  0.046, SRMR =  0.028 for the 12-item RRI-C. McDonald’s omega total was 0.97 and 0.95 respectively. The correlation between the total scores of the two versions was r =  0.98; the average discrepancy was 1.85 points higher for the comprehensive version with a slope of -0.013 (p =  0.12). Both versions showed functionally identical reliability and factor structure when therapy is online vs. in-person. Significant correlations were found between the RRI-C and the Working Alliance Inventory (r =  0.68 and r =  0.67 for the 24-item and 12-item versions, respectively, both p < .001) and the Session Evaluation Scale (r =  0.62 and r =  0.58, respectively, both p <  0.001). This study substantiates the sound psychometric properties of the 24-item and 12-item RRI-C.