Category: Transdiagnostic

PS6- #A16 - Hidden Problems in the Measurement of Experiential Avoidance: Indiscrimination and Tautologies

Friday, Nov 17
2:45 PM – 3:45 PM
Location: Indigo Ballroom CDGH, Level 2, Indigo Level

Keywords: Experiential Avoidance | ACT (Acceptance & Commitment Therapy)

Experiential avoidance reflects the unwillingness to experience uncomfortable thoughts, feelings, and bodily sensations - thereby moving beyond the behavioral avoidance of situations to also include internal states (Hayes, Wilson, Gifford, Follette, & Strosahl, 1996). Despite being one of the most studied transdiagnostic constructs in the past decade, measurement issues exist. Since experiential avoidance captures beliefs and behaviors in response to distress (i.e., meta-emotions), existing self-report measures run the risk of simply assessing distress rather than reactions.


We compared two self-report measures of experiential avoidance:  The Acceptance and Action Questionnaire (AAQ-II) and the Brief Multidimensional Experiential Avoidance Questionnaire (BMEAQ). We hypothesized that the meta-emotional items of the AAQ-II and BMEAQ would relate strongly to psychological distress such that items would lack discriminant validity. To test our hypotheses, we conducted an online survey study with American adults. We recruited 403 participants via Amazon’s Mechanical Turk (July, 2016). With respect to psychological distress, the Mini International Personality Item Pool measured neuroticism, the Brief Mood Induction Measure measured negative affect, and the Depression, Anxiety, and Stress Scales measured internalizing symptoms.


We found that all 7 AAQ-II items correlated above .50 with neuroticism and negative affect total scores and correlated above .60 with depression, anxiety, and stress total scores. Of the 16 BMEAQ items, more than half correlated below .25 with neuroticism, negative affect, depression, anxiety, and stress total scores; all other items correlated below .40. We then tested whether the AAQ-II and BMEAQ represented a factor different from psychological distress. Using Mplus 7.4, total scores for the AAQ-II and BMEAQ were correlated with a latent factor indicated by neuroticism, negative affect, and internalizing symptoms. The AAQ-II correlation was .85 while the BMEAQ correlation was .50. The BMEAQ items and total score appears to have greater discriminant validity from psychological distress than the AAQ-II.


The high associations between the AAQ-II and psychological distress suggest the AAQ-II could arguably be conceptualized as a measure of psychological distress rather than experiential avoidance (Wolgast, 2014). This is problematic for using the AAQ-II as a predictor of distress-related outcomes or changes in distress following interventions. In future research, we suggest the BMEAQ be used to measure experiential avoidance rather than the AAQ-II. Our results highlight the importance of discriminant validity when measuring meta-emotions to prevent inflated effects due to tautological item content

David J. Disabato

Clinical Doctoral Student
George Mason University

Todd B. Kashdan

Full Professor
George Mason University

Fallon R. Goodman

Clinical Doctoral Student
George Mason University

Maria A. Larrazabal

Post-Baccalaureate
George Mason University

John T. West

Post-Baccalaureate
George Mason University