Enhancing tolerability of a measure of social perception in schizophrenia: comparison of short and long Norwegian versions of the Relationships Across Domains test.

INTRODUCTION

Relationship perception focuses on social interactions, is reduced in schizophrenia and is related to daily functioning. It can be assessed with the Relationships Across Domains (RAD) test, built on Relational Models Theory which states that people use four relational models to interpret social interaction. RAD is time consuming, low on tolerability and only used in English-speaking countries. We evaluated the psychometric properties of a translated, abbreviated Norwegian version.

METHODS

Sixty-two schizophrenia participants and 56 healthy controls underwent assessments of social and non-social cognition. The schizophrenia group completed functional and clinical measures. RAD's internal consistency was investigated with Cronbach's alphas, group differences with logistic regressions and associations between study variables with Pearson's correlations.

RESULTS

RAD was reduced from 25 (Cronbach's alpha = .809) to 12 vignettes (Cronbach's alpha = .815). Schizophrenia participants had significant impairments, with larger effect sizes for the full version. Associations of RAD with study variables were similar for the two versions: smaller for clinical measures and larger for functional and cognitive measures. Results were comparable to results for the English version.

CONCLUSIONS

The length of the Norwegian RAD was reduced while retaining its psychometric properties, which were similar to the English version. This suggests the test's cross-cultural utility.