Craig A Velozo1, Mary Warren, Eric Hicks, Kathleen A Berger. 1. North Florida/South Georgia Veterans Health System, Research Service, Department of Veterans Affairs Medical Center, College of Public Health and Health Professions, University of Florida, Gainesville, Florida, USA. velozo@musc.edu
Abstract
PURPOSE: Visual functioning questionnaires (VFQs) address patient-centered and evidenced-based practice. The purpose of this study was to investigate the psychometrics of the Self-Report Assessment of Functional Vision Performance (SRAFVP) and present Rasch-generated patient-level outputs. METHODS: One hundred two patients were administered the 39-item SRAFVP that includes questions under the categories of reading, writing, money management, telephone usage, reading a timepiece, personal care, clothing care, meal preparation, leisure, and functional mobility. Dimensionality was determined using confirmatory factor analysis and principal components analysis of residuals. The Andrich rating (Rasch) model was used to determine rating scale qualities, item/person fit to the Rasch model, item-difficulty hierarchy, person-item match, and person separation. Keyform outputs were generated that present response patterns for individual patients. RESULTS: Whereas the confirmatory factor analysis confirmed reading and mobility dimensions, the principal components analysis indicated that the SRAFVP formed a single dominant measure. The original five-point rating scale showed disordered thresholds that were corrected by collapsing the three middle rating categories. Only two of 39 items showed high infit and outfit statistics and five of 39 items showed high outfit statistics. The items showed a logical item-difficulty order related to the visual ability necessary to accomplish activities of daily living. The SRAVFP separated the sample into 3.77 distinct strata, showing no ceiling or floor effects. Keyforms demonstrated distinct patterns of responses for individuals reporting different visual functioning ability. CONCLUSIONS: As with many VFQs, the SRAVFP shows good item-level psychometric qualities. The SRAVFP and other VFQs may provide the foundation for theory-based instrument calibration that should enhance our translation of measures across VFQs and improve our understanding of visual functioning. Rasch measurement provides a means to generate outputs that may prove to be useful in goal setting and treatment planning when providing low-vision rehabilitation.
PURPOSE: Visual functioning questionnaires (VFQs) address patient-centered and evidenced-based practice. The purpose of this study was to investigate the psychometrics of the Self-Report Assessment of Functional Vision Performance (SRAFVP) and present Rasch-generated patient-level outputs. METHODS: One hundred two patients were administered the 39-item SRAFVP that includes questions under the categories of reading, writing, money management, telephone usage, reading a timepiece, personal care, clothing care, meal preparation, leisure, and functional mobility. Dimensionality was determined using confirmatory factor analysis and principal components analysis of residuals. The Andrich rating (Rasch) model was used to determine rating scale qualities, item/person fit to the Rasch model, item-difficulty hierarchy, person-item match, and person separation. Keyform outputs were generated that present response patterns for individual patients. RESULTS: Whereas the confirmatory factor analysis confirmed reading and mobility dimensions, the principal components analysis indicated that the SRAFVP formed a single dominant measure. The original five-point rating scale showed disordered thresholds that were corrected by collapsing the three middle rating categories. Only two of 39 items showed high infit and outfit statistics and five of 39 items showed high outfit statistics. The items showed a logical item-difficulty order related to the visual ability necessary to accomplish activities of daily living. The SRAVFP separated the sample into 3.77 distinct strata, showing no ceiling or floor effects. Keyforms demonstrated distinct patterns of responses for individuals reporting different visual functioning ability. CONCLUSIONS: As with many VFQs, the SRAVFP shows good item-level psychometric qualities. The SRAVFP and other VFQs may provide the foundation for theory-based instrument calibration that should enhance our translation of measures across VFQs and improve our understanding of visual functioning. Rasch measurement provides a means to generate outputs that may prove to be useful in goal setting and treatment planning when providing low-vision rehabilitation.
Authors: Tiffany L Chan; Monica S Perlmutter; Melva Andrews; Janet S Sunness; Judith E Goldstein; Robert W Massof Journal: Arch Phys Med Rehabil Date: 2015-07-09 Impact factor: 3.966