Text mining a self-report back-translation.

Area of Science:

• Psychology
• Computational Linguistics
• Cross-Cultural Research

Background:

• Standard back-translation procedures for cross-cultural research instruments lack systematic evaluation.
• Text mining methods have been underutilized in adapting self-report instruments across languages.

Purpose of the Study:

• To introduce an innovative text mining application for adapting personality questionnaires into 12 languages.
• To enhance the accuracy and systematic evaluation of the back-translation process in cross-cultural research.

Main Methods:

• A three-stage text mining method: descriptive analysis of back-translations, dissimilarity assessment between source and back-translated versions, and item meaning equivalence evaluation.
• Application to a personality questionnaire adapted into 12 different languages.

Main Results:

• The proposed method provides a systematic approach to evaluating the equivalence of instrument versions across languages.
• It leads to more accurate instrument back-translations, positively impacting reliability and validity in diverse cultural contexts.

Conclusions:

• The text mining approach offers a structured and informed evaluation for cross-cultural instrument adaptation.
• This methodology can improve the reliability and validity of test scores across different languages and cultures, with potential applications in clinical assessment.