THE RELATIONSHIP ASSESSMENT SCALE REVISED (RAS-R): ADAPTATION AND PSYCHOMETRIC EVALUATION BY CLASSICAL TEST THEORY AND ITEM RESPONSE THEORY

The study introduces the revised Relationship Assessment Scale (RAS-R) in response to discrepancies identified within existing RAS models. The first phase involved grammatical revisions, bias removal, and standardization of response options for the Spanish version of the RAS. The subsequent phase evaluated the psychometric properties of both the RAS and the RAS-R, as well as the specification of com-mon models from the literature, using data from 547 participants for the RAS and 285 for the RAS-R. In contrast to the RAS, which required adjustments due to correlated error or item elimination, the RAS-R demonstrated optimal fit without requiring model respecification. Further analysis using the Graded Response Model (GRM) indicated a better fit for the RAS-R, with consistent item discrimination across items, particularly for Item 2. Evaluation of test and item curves revealed a wide range of responses, highlighting the reliability of the instrument. In conclusion, the RAS-R was found to be a robust instrument that addresses the factorial structure challenges encountered in the RAS literature.