Psychometric evaluation of interpretation bias measures for pain

Negative interpretation bias, the tendency to interpret ambiguous information in a threatening way, is thought to play an important role in the experience and maintenance of chronic pain. However, evidence for valid and reliable measures of interpretation bias in pain remains limited. This study examined the psychometric properties of multiple interpretation bias measures in individuals with chronic pain and healthy controls. Participants (N=451; N=235 chronic pain, N=188 healthy) completed several established measures of interpretation bias twice with two weeks in between, including the Ambiguous Sentences (AS) Task, the Encoding Recognition (ER) task, the Bodily Sensations Interpretation Questionnaire (BSIQ), and the Word Association Task (WAT), alongside outcomes of pain, negative response bias, and social desirability. Analyses were conducted to evaluate their validity including convergent, known-groups, structural, criterion and discriminant validity. Reliability was assessed through internal consistency and test-retest reliability. The tasks appear to capture a shared underlying construct and overall demonstrate good reliability within conventions appropriate for implicit cognitive measures. Most measures showed modest but significant associations with pain outcomes, and group differences emerged for the BSIQ and WAT. Most tasks were not affected by a general negative response bias, with the exception of the AS task. Findings provide preliminary support for the use of the BSIQ and WAT as reliable tools for assessing pain-related interpretation bias. PERSPECTIVE: This study evaluates the psychometric properties of commonly used language-based interpretation bias tasks for pain. By identifying measures with the strongest psychometric properties, this work supports more rigorous assessment of cognitive mechanisms implicated in pain and may guide future clinical and experimental research.