Measuring Psychological (In)flexiblity: A Psychometric Evaluation and Comparison of Existing Measures

Abstract

Psychological (in)flexibility, the tendency for behavior to be guided by psychological reactions rather than chosen values, is a central construct of interest in acceptance and commitment therapy (ACT; Hayes, Strosahl, & Wilson, 2012; Levin et al., 2014). Psychological (in)flexibility has received significant attention as a potentially important mechanism for understanding the development and maintenance of psychopathology. Although several self-report measures of this construct have been developed, some of the most widely used measures have been identified as having potentially serious problems. Additionally, there is a lack of consensus regarding which of these measures is most theoretically and psychometrically sound. In fact, the majority of recently developed measures of psychological (in)flexibility have not been independently evaluated. The overarching aim of the current study was to examine the reliability and validity of each of the identified measures of psychological (in)flexibility and provide further information about sensitivity of the measures to change over time. A sample of 474 adults were included in the sample at Time 1 and 127 participants were retained for a second assessment at Time 2. In general, results of reliability analyses suggested adequate internal consistency for most of the measures of psychological (in)flexibility that were examined in this study, though item redundancy emerged as an area of concern for several measures. Results of contrast analysis provided evidence that the Acceptance and Action Questionnaire-II (AAQ-II), Acceptance and Action Questionnaire-3 (AAQ-3), and Multidimensional Psychological Flexibility Inventory (MPFI) Inflexibility scale demonstrated the best fit within the specified nomological network. Despite these measures providing the best fit within the nomological network, their discriminant validity was questionable. Finally, results of sensitivity to change analysis indicated that most of the target measures were not particularly sensitive to changes in psychological (in)flexibility over time. Findings from this study represent a step forward in refining the measurement of psychological (in)flexibility, and, in turn, improving understanding of the construct more generally.