Use of item response analysis to investigate measurement properties and clinical validity of data for the dynamic gait index.

BACKGROUND AND PURPOSE The Dynamic Gait Index (DGI) is a standardized clinical assessment that aids in evaluating a subject's ability to modify gait in response to changing demands. The purpose of this study was to use Rasch measurement theory to examine whether the DGI rating scale meets suggested psychometric guidelines, whether the hierarchical order of DGI tasks is consistent with a clinically logical testing procedure, and whether the DGI represents a unidimensional construct. SUBJECTS Subjects were 84 community-dwelling male veterans (age range=64-88 years; mean+/-SD=75+/-6.47 years). METHODS Data were retrieved retrospectively from the participants' clinical records. The Rasch measurement model with the WINSTEPS program was used in this study because it offers distinct advantages over traditional psychometric approaches. RESULTS Overall, the DGI showed sound item psychometric properties. Each of the original 4 rating scale categories appeared to distinctly identify subjects at different ability levels. The analysis revealed a clear item difficulty hierarchical order that is generally consistent with clinical expectations. In addition, fit statistics and principal components analysis indicated that the 8 items of the DGI appear to represent a single construct. DISCUSSION AND CONCLUSION The results suggest that the rating scale of the DGI is used appropriately for community-dwelling older subjects with balance problems. The findings support the continued use of this well-constructed scale for clinical and research assessment in a community-dwelling population of older subjects.