Does the Brief-BESTest Meet Classical Test Theory and Rasch Analysis Requirements for Balance Assessment in People With Neurological Disorders?

摘要: Background The Brief Balance Evaluation System Test (Brief-BESTest) has been recently proposed as a clinical tool for quickly measuring balance disorders, but its measurement properties require investigation. Objective To perform detailed analysis of the psychometric Brief-BESTest, by means Classical Theory and Rasch Analysis. Design Observational study. Methods Brief-BESTest data were collected in sample 244 patients. Internal consistency was analyzed Cronbach's α item-to-total correlation. Test-retest inter-rater reliability investigated subgroup 21 Minimum Detectable Change calculated at 95% confidence level (MDC95). Scale dimensionality examined through Horn's Parallel Analysis followed exploratory factor ordinal data. Finally, underwent (rating scale model). Results 0.89 all correlations > 0.40. showed an ICC(2,1) = 0.94 0.90. MDC95 equal to 4.30 points. test's unidimensionality confirmed one item low communality. inadequacy response categories, 5 misfitting items, minor mistargeting, moderate person (0.80), 2 pairs locally dependent items. Limitations Our cross-section subjects with disorders from different neurological etiologies consecutively recruited single rehabilitation facility. Conclusions presence some acceptable good indices when according Theory, quite limited sensitivity change. indicated that selection should be improved point view. There is need reduce redundancy improve metric coverage measured construct new items.