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

Shantini Oorjitham,Oleg N. Medvedev,Adrian J. Bravo,Christopher C. Conway,James M. Henson,Lee Hogarth,Manuel I. Ibáñez,Debra Kaminer,Matthew Keough,Laura Mezquita,Generós Ortet,Matthew R. Pearson,Angelina Pilatti,Mark A. Prince,Jennifer P. Read,Hendrik G. Roozen,Paul Ruiz,Cross-Cultural Addictions Study Team
DOI: https://doi.org/10.1007/s12144-024-05894-w
IF: 2.8
2024-04-10
Current Psychology
Abstract:The 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.
psychology, multidisciplinary
What problem does this paper attempt to address?
The main problems that this paper attempts to solve include: 1. **Limitations of Ordinal Response Format**: The Distress Tolerance Scale (DTS) is a widely - used psychometric tool, but its ordinal response format (such as the 5 - point Likert scale) has limitations. These limitations include: - The distances between response options are not equal, and it is impossible to quantify the degree of participants' agreement with items. - Analyzing ordinal data using parametric statistical methods violates basic assumptions and may lead to invalid results. - The method of calculating the total score by simply adding up the scores of each item is not precise enough because each item contributes differently to the underlying construct. 2. **Validity of Cross - Cultural Comparison**: Although the DTS has been translated into multiple languages and used worldwide, current research has not fully verified its cross - cultural validity in different countries and language versions. Specifically: - People in different countries may respond to DTS items in different ways, which will affect cross - cultural comparisons. - Translated scales may have problems such as non - equivalent content or different levels of difficulty in understanding, further affecting the accuracy of cross - cultural comparisons. 3. **Multidimensional Structure vs. Unidimensionality Controversy**: The DTS was initially designed to include four sub - scales (Tolerance, Absorption, Appraisal, Regulation), but some studies have shown a high correlation among these sub - scales, suggesting that the DTS may be more appropriately regarded as a unidimensional construct. Therefore, more evidence is needed to support the unidimensionality of the DTS and the validity of its total score. 4. **Measurement Invariance**: To ensure the consistency and validity of the DTS in different populations (such as gender, meditation practice, country), its measurement invariance (Measurement Invariance) must be evaluated. In particular, it is necessary to check for differential item functioning (DIF), that is, whether different groups with the same DT level respond consistently to the same item. ### Solutions To solve the above problems, the author adopted the Rasch analysis method with the following specific goals: - **Evaluating Unidimensionality**: Evaluate the unidimensionality of the DTS through Rasch model fit statistics and principal component analysis of residuals (PCAR). - **Evaluating Measurement Invariance**: Evaluate the measurement invariance of the DTS among genders, meditation practices, and countries through DIF analysis. - **Evaluating Internal Consistency**: Evaluate the internal consistency of the DTS from a perspective different from classical test theory, using the person separation index (PSI). - **Improving Rasch Model Fit**: If the initial fit is not good, improve the model fit by deleting items or combining locally dependent items into testlets, and develop an algorithm for conversion from ordinal to interval to improve the precision of DT measurement. Finally, this study aims to improve the psychometric properties of the DTS so that it can be more accurately used in research and clinical applications for people from different countries and language backgrounds.