Objectives: To test the validity and reliability of measures of visual ability and to evaluate the relation between measurements made at the task level and measurements made at the goal level of a hierarchical model for visual disability.
Design: Validation of a telephone-administered functional assessment instrument using Rasch analysis on self-assessment ratings.
Setting: Telephone interviews of respondents in their homes.
Participants: Consecutive series of 600 outpatients with low vision.
Interventions: Not applicable.
Main outcome measures: Ordinal ratings of the difficulty in performing a subset of 337 tasks. Measures of the visual ability of each patient and the required visual ability to perform each task were made using the Andrich rating scale model. Measurement validity and reliability were tested statistically by comparing response patterns and distributions to measurement model expectations.
Results: Results were consistent with a single visual ability construct. Patients' visual ability estimated from task difficulty ratings agreed with estimates from goal difficulty ratings ( r =.74); the difficulty of individual goals was equal to the weighted average of the difficulties of subsidiary tasks ( r =.79). However, conclusions from the Rasch analysis were not confirmed by principal components analysis of item residuals, which indicated that visual ability had a 2-dimensional structure, with 1 factor related to mobility and the other related to reading. Factor analysis on person measures estimated from subsets of functionally grouped items confirmed the 2-dimensional structure of visual ability.
Conclusions: Our study results confirm the hierarchical structure of the Activity Breakdown Structure model and show how the individualized Activity Inventory can produce measures of limitations in functional vision.