Respondent impact on functional ability outcome measures in vision rehabilitation

To assess the relative importance of several measurement facets including respondent (patient or clinicians), item (functional domains), and administration (pre- and postrehabilitation) on assessing perceived functional outcomes in vision rehabilitation.

A convenience sample of VA veterans were assessed with a self-report and clinician rated perceived functional ability instrument before and after rehabilitation services. The items were analyzed using a Rasch model to ensure that both versions fit the restrictive model well enough to be directly compared. A second analysis using a generalizability theory model assessed the relative importance of each measurement facet.

The Rasch model supported the equivalence of the two different versions of the outcome instrument. The generalizability theory model showed that administration accounted for the majority of the variance and that respondents and items contributed very little to perceived functional assessment.

The results support the conclusion that both clinical ratings and patient self-reported perceived functional ability ratings provide relative equivalent values in blind and low-vision rehabilitation settings. The relative importance of administration time in predicting scores may support the general conclusion that rehabilitation produces the largest relative changes in perceived functional ability and ought to shift researchers' focus away from the subtleties of instrument development and more toward assessing program and individual outcomes. Finally, these findings suggest that there is little promise in modeling individual differences to account for perceived functional ability changes because of rehabilitation.