Cross-national examination of the Distress Tolerance Scale using Rasch methodology

Abstract

AbstractThe 15-item Distress Tolerance Scale (DTS) is a widely used psychometric measure with evidence generally supporting its reliability and validity. However, limitations related to its ordinal response format and lack of cross-cultural comparisons have yet to be investigated using appropriate methods. The Partial Credit Rasch model was used to evaluate and enhance the psychometric properties of the DTS using responses from 2550 adult participants from the United States of America (USA), England, Canada, South Africa, Spain, and Argentina. The initial poor fit of the DTS to the Rasch model was improved by removing one item and combining locally dependent items into three testlets. These modifications resulted in the best fit of the 14-item DTS to the Rasch model for all the countries in our study, providing evidence of unidimensionality, high reliability and invariance across countries, meditation practice, and gender. Meeting the expectations of the Rasch model permitted the development of ordinal-to-interval conversion algorithms derived from person estimates of the Rasch model. Using the ordinal-to-interval conversion algorithms published in this article, ordinal DTS scores can be transformed into interval-level data, enhancing the precision of this scale for future research and clinical use across people from the six countries in this study and across the English and Spanish versions of the 14-item DTS.