Enhanced Interpretation of Instrument Scales Using the Rasch Model
- PDF / 100,160 Bytes
- 10 Pages / 612 x 792 pts (letter) Page_size
- 8 Downloads / 228 Views
John R. Thompson Pfizer Global Research and Development, New London, Connecticut Joseph C. Cappelleri Pfizer Global Research and Development, Groton, Connecticut Christine Getter Pfizer Global Research and Development, New London, Connecticut Andreas Pleil Pfizer Global Research and Development, La Jolla, California Martin Reichel Universitatsaugenklink Leipzig, Leipzig, Germany Sebastian Wolf Klinik und Poliklinik fuer Augenheilkunde, Inselspital, University Bern, Bern, Switzerland
Key Words Instrument scales; Item response theory; Age-related macular degeneration; Rasch model; Visual function questionnaire Correspondence Address John R. Thompson, Pfizer Global Research and Development, 50 Pequot Avenue, MS MS6025-A4168, New London, CT 06320 (e-mail: [email protected]).
541
Enhanced Interpretation of Instrument Scales Using the Rasch Model Objectives: To enhance the interpretability of instrument scales in general and the National Eye Institute Visual Function Questionnaire in particular. Methods: The Rasch model was applied to the German version of the 39-item National Eye Institute Visual Function Questionnaire in patients with age-related macular degeneration. Two important subscales, near vision and distance vision, were used in this work. The likelihood of performing a task with little or no difficulty was selected as an indicator of a patient’s quality of life. Each possible subscale score was related to the likelihood of performing each task constituting the subscale with little or no difficulty. Results: The Rasch rating scale model was found to fit both subscales based on acceptable infit mean
INTRODUCTION Multiple instruments have been developed to measure vision-related quality of life, including the Visual Functioning Index (VFI) (1), the Visual Activities Questionnaire (VAQ) (2), the Activities of Daily Vision Scale (ADVS) (3), the Visual Performance Questionnaire (VPQ) (4), the 14item Visual Functioning Index (VF-14) (5), the Visual Disability Assessment (VDA) (6), the Low Vision Quality of Life Questionnaire (LVQOLQ) (7), and the National Eye Institute Visual Function Questionnaire (NEI-VFQ) (8,9). All of these instruments result in a score that can be difficult to interpret. The objective of this work was to develop an alternative presentation of the results that would add interpretability to instrument results in general and the NEI-VFQ in particular. The National Eye Institute sponsored the development of the NEI-VFQ to assess the dimensions of vision-targeted functional health status that are most important to persons with chronic eye diseases (8,9). The NEI-VFQ comes in a long form with 51 items (8) and a short form with 25
squares, a wide range of item difficulties, and monotonically increasing thresholds. Rasch category characteristic curves for each item in the subscale were back-transformed to the original scale of measurement and placed on a single graph. The graph provides a depiction of the relative difficulty of tasks, the probability of an individual performing a task wit
Data Loading...
