Measurement validity of the low vision independence measure (LVIM)

OBJECTIVE. This study's objective was to test the psychometrics of Smith's (2013) Low Vision Independence Measure (LVIM) using the Rasch model. METHOD. A cohort design was used with 93 participants receiving occupational therapy for low vision. Confirmatory factor analysis (CFA) with a parceling approach was used to test the LVIM factors and Rasch analysis to examine item-level psychometrics. RESULTS. Participants' average age was 78.9 yr (standard deviation = 12.1), and the majority were female (72.8%) with macular degeneration (62.3%). The CFA revealed two measurement factors: visual field or scotoma (n = 28) and visual acuity (n = 24). We removed six misfitting items, and the two factors of the revised LVIM demonstrated good rating scale function, good internal consistency (person reliability: visual field, .87; visual acuity, .90), good precision (person strata: visual field, 3.91; visual acuity, 4.40), no ceiling or floor effects, and no differential item functioning. CONCLUSION. The revised LVIMdemonstrates good psychometrics on the Rasch model and can be used as a valid outcome measure in low vision rehabilitation.