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

Quispilaya-Capcha K.L.; Yupanqui-Lorenzo D.E.; Montes-Herrera G.; Romero-Alvarez F.M.; Arauco-Lozada T. · TPM - Testing, Psychometrics, Methodology in Applied Psychology (2024) · DOI: 10.4473/tpm31.3.1

Resumen

The study introduces the revised Relationship Assessment Scale (RAS-R) in response to discrepan-cies 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 eval-uated the psychometric properties of both the RAS and the RAS-R, as well as the specification of common 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.

Instituciones: Universidad Peruana Cayetano Heredia; Universidad Nacional Federico Villarreal; Universidad Privada del Norte; Universidad de Ciencias y Humanidades