The examination of multidimensional psychological flexibility inventory: Application of factor analysis and item response theory

Abstract

Background and Objective: The Multidimensional Psychological Flexibility Inventory (MPFI; Rolffs et al., 2018) is a recently developed self-report measure of psychological flexibility and psychological inflexibility. Using factor analyses and an item response theory (IRT) approach, my dissertation project aimed to examine and validate the MPFI with a racially and ethnically diverse sample of adults in Hawaiʻi, and to develop and validate a new shorter-version of MPFI (MPFI-S). Method: The present dissertation project consists of two phases (i.e., Study 1 and Study 2) with two convenience samples of racially diverse participants recruited from undergraduate psychology courses at the University of Hawaiʻi at Mānoa. The first phase (Study 1: n = 719) had five specific foci. First, the factor structure of the MPFI was examined through exploratory factor analysis (EFA) and confirmatory factory analysis (CFA). Second, following the framework of IRT, the functions of each item across purported trait levels were investigated. Third, the study assessed whether the original MPFI maintains measurement invariance at the structural level across gender and race/ethnicity groups. Fourth, preliminary validity evidence was gathered. Fifth and finally, a shorter version of the MPFI (i.e., MPFI-S) was preliminarily proposed by identifying three items for each dimension of psychological flexibility and psychological inflexibility, respectively. Following the completion of phase 1, Study 2 with another sample (n = 698) examined and psychometrically validated the short version of MPFI (MPFI-S). More specifically, the factor structure of MPFI-S was compared to that of MPFI derived from Study 1, and item functions were subsequently analyzed with the IRT application. Finally, the study preliminarily investigated the convergent validity as well as divergent validity of the specific domains of psychological flexibility and inflexibility measured by the MPFI-S. Results: The findings of Study 1 support the factor structure of MPFI that was confirmed by the original MPFI psychometric study (Rolffs et al., 2018). More specifically, the factor structure found in Study 1 consisted of two higher-order factors, each having six respective first-order factors. Additionally, measurement invariance was found for gender and race/ethnicity categories, and IRT analysis indicated that most items demonstrated optimal functioning at the item level. Furthermore, generally, correlational findings supported the construct validity of MPFI scores. Finally, based on these findings and a review of item contents, a 36-item MPFI-S was preliminarily developed. Study 2 results revealed that the MPFI-S is a theoretically and psychometrically sound measure, which is comparable to the original MPFI. More specifically, Study 2 confirmed the presence of two higher-order factors in the MPFI-S, each having six corresponding first-order factors. IRT analyses also indicated that all items in the MPFI-S and their categories functioned optimally. Finally, correlational findings preliminarily supported the construct validity of MPFI-S scores at the first-order factor levels. Discussion: With Hawaiʻi-based samples of racially diverse college students, my dissertation project offers preliminary evidence of the MPFI and MPFI-S as psychometrically sound self-report measures of psychological flexibility and inflexibility.