Psychometric properties of the Oral Behaviors Checklist: Preliminary findings

Objective: The assessment of oral parafunctional behaviors, as one source of muscle overuse possibly related to TMD, remains complex due to inaccuracies of self-report for these largely unconscious behaviors. We developed the Oral Behaviors Checklist (OBC) as one possible assessment tool. Content validity emerged from both expert as well as patient contributions. The present study investigates basic psychometric properties in a sample of subjects selected for simultaneous laboratory investigation (Markiewicz et al, 2004). Methods: The 23-item questionnaire (rated 0-4, from “never” to “all of the time”) was administered at the onset of treatment for TMD cases (n=13), and then during the subsequent laboratory session for these 13 cases and 11 controls. Analyses included internal reliability, test-retest reliability, and comparison of means. Results: TMD subjects report more frequent behaviors (p=.0004) across multiple items. Internal reliability was high (alpha=0.86) for 16 items whose content is consistent with common clinical observations. Reported behavior frequency at treatment onset correlated -0.17 with remembered frequency obtained in the laboratory, though mean scores at these two intervals were not significantly different. Sensitivity to change indicated 25% reported reduction in behavior frequency over a moderately short period of time during which treatment was provided, but this reduction in reported OBC frequency related poorly to global change in frequency of behaviors. Conclusion: Within the constraints imposed by the target behaviors occurring largely outside of conscious awareness and a small at present sample size, these data demonstrate that a checklist for such behaviors can have content validity, excellent internal reliability, at least simple construct validity, and possible utility at measuring outcomes. Further investigation of this challenging construct will require, in addition to more subjects, attention to sufficient level of awareness such that high test-retest reliability of the measurement tool can be obtained. Research supported by NIH DE-13221 and T-35 DE-07106.