Item Response Theory Analysis of the Five Facet Mindfulness Questionnaire and its Short Forms.

Abstract

The Five Facet Mindfulness Questionnaire (FFMQ) is a self-report measure of mindfulness with forms of several different lengths, including the FFMQ-39, FFMQ-24, and FFMQ-15. We use item response theory analysis to directly compare the functioning of these three forms. Data were drawn from a non-clinical Amazon Mechanical Turk study (N = 522) and studies of aftercare treatment of individuals with substance use disorders (combined N = 454). The item and test functioning of the three FFMQ forms were studied and compared. All 39 items were strongly related to the facet latent variables, and the items discriminated over a similar range of the latent mindfulness constructs. Items provided more information in the low to medium range of latent mindfulness than in the high range. Scores in three of the five FFMQ-39 facets were unreliable when measuring individuals in the high range of latent mindfulness, resulting from ceiling effects in item responses. Reliability in the high range of mindfulness was further reduced in the FFMQ-24 and FFMQ-15, such that short forms may be ill-suited for applications that require reliable measurement in the high range. Results suggest the existing FFMQ item pool cannot be reduced without negatively affecting either overall reliability or the span of mindfulness over which reliability is assessed. Conditional test reliability curves and item functioning parameters can aid investigators in tailoring their choice of FFMQ form to the reliability they hope to achieve and to the range of latent mindfulness over which they must reliably measure.