Translation of psychological measures into multiple languages increases their accessibility for researchers and clinicians working with linguistically diverse populations. However, the validity and utility of translated instruments depends on the quality of their translation, and differences across translations in key linguistic characteristics, such as sentiment and emotional tone/intensity, could introduce unintended but systematic measurement bias. This proof-of-concept study introduces and demonstrates the use of natural language processing (NLP), a subfield of artificial intelligence, to linguistically analyze 13 translations of two personality measures that have been previously translated into numerous languages. NLP was used to generate estimates reflecting specific linguistic characteristics of test items (i.e., emotional tone/intensity, sentiment, valence, arousal, and dominance), which were then compared across translations at both the test- and item-level, as well as between the two instruments. Following a discussion of results, I summarize the current limitations of this approach and propose strategies for advancing this methodology.
No takes yet. Share an insight, caveat, or question.
Adam P. Natoli (2024) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: