Functional movement screening (FMS) is a musculoskeletal assessment that is intended to fill a gap between preparticipation examinations and performance tests. Functional movement screening consists of 7 standardized movements involving multiple muscle groups that are rated 0-3 during performance; scores are combined into a final score, which is intended to predict injury risk. This use of a sum-score in this manner assumes that the items are unidimensional and scores are internally consistent, which are measures of internal reliability. Despite research into the FMS' predictive value and interrater reliability, research has not assessed its psychometric properties. The present study is a standard psychometric analysis of the FMS and is the first to assess the internal consistency and factor structure of the FMS, using Cronbach's alpha and exploratory factor analysis (EFA). Using a cohort of 877 male and 57 female Marine officer candidates who performed the FMS, EFA of polychoric correlations with varimax rotation was conducted to explore the structure of the FMS. Tests were repeated on the original scores, which integrated feelings of pain during movement (0-3), and then on scores discounting the pain instruction and based only on the performance (1-3), to determine whether pain ratings affected the factor structure. The average FMS score was 16.7 ± 1.8. Cronbach's alpha was 0.39. Exploratory factor analysis availed 2 components accounting for 21 and 17% and consisting of separate individual movements (shoulder mobility and deep squat, respectively). Analysis on scores discounting pain showed similar results. The factor structures were not interpretable, and the low Cronbach's alpha suggests a lack of internal consistency in FMS sum scores. Results do not offer support for validity of the FMS sum score as a unidimensional construct. In the absence of additional psychometric research, caution is warranted when using the FMS sum score.

The present study extends the item-level psychometric information of the Leeds Dependence Questionnaire (LDQ; Raistrick et al., 1994) that has been purported to measure psychological dependence and the International Statistical Classification of Diseases and Related Health Problems-10th edition substance dependence criteria. Prior research on the LDQ has not established item-level properties or the degree of differential item functioning (DIF) by gender and substance type. Principal component and Mokken scale analyses were used to assess unidimensionality and monotonicity of the responses to the scale items, respectively. Graphical and statistical methods examined the model-data fit of the graded response model and two-parameter logistic model of LDQ responses (n = 1,681) obtained from 2 community treatment centers. DIF analysis was performed on gender (men = 1,313, women = 353) and substance (alcohol = 821, opiates = 528) groups. The 2PL achieved the best model-data fit. Three items provided little information about standing on the underlying construct, indicating that they are likely not good indicators of the "pure" psychological construct the LDQ aims to measure. Overall, the LDQ offers the greatest precision in quantifying psychological dependence in a clinical sample along the lower to mid ranges of this construct. Uniform DIF was present in Item 7 of the dichotomized responses by substance (alcohol vs. opiates). DIF by gender was not found in any of the LDQ items. Recommendations include revising the scaling and discussing the need to obtain LDQ data from different levels of care and primary identified substance. (PsycINFO Database Record

No abstract

No abstract