LAUSR

Abstract This paper describes the development and initial support for Inflexitext, an automated program identifying psychological inflexibility in unstructured verbal data. Written in Python 3.7, Inflexitext produces a psychological inflexibility score based on patterns of word occurrence reflecting its contributing processes. Inflexitext performance was examined in a sample of 809 English speaking adults in the United States recruited using Amazon's Mechanical Turk platform. Participants wrote essays in response to a prompt to write about an emotional issue and completed self-report measures of distress and psychological flexibility relevant constructs. Participant essays were analyzed using Inflexitext and Linguistic Inquiry Word Count 2015 (LIWC), a popular text scoring program. Inflexitext scores demonstrated small positive correlations to self-report measures of experiential avoidance, cognitive fusion, challenges in progress towards one's values, and to symptoms of depression, anxiety, and stress and a medium positive correlation with LIWC coding of negative emotion. Inflexitext scores evidenced small negative correlations with progress towards one's values and LIWC scores on positive emotion. Overall, this initial examination provides preliminary support for the program, although further evaluation is needed and limitations are discussed. Potential applications for future development include unobtrusive ambient monitoring of verbal behavior and real time examination of psychological inflexibility as related to psychological functioning and therapeutic outcomes.