Parsimony and Model Confirmation Of the ADL Self-Care Scale For Multiple Sclerosis Persons

Data from a 52-item self-administered ADL Self-Care Scale designed for persons diagnosed with multiple sclerosis were refactored for the purpose of achieving scale parsimony and clarifying interrelationships among ADL self-care behaviors.

Abstract

Data from a 52-item self-administered Activities of Daily Living (ADL) Self-Care Scale designed for persons diagnosed with multiple sclerosis (MS) were refactored for the purpose of achieving scale parsimony and clarifying interrelationships among ADL self-care behaviors. Analyses were based on 629 MS subjects. Principal component analysis and Varimax rotated factor analysis were used to achieve scale parsimony. The resultant 15-item scale was subjected to confirmatory factor analysis to determine relationships among factors, observed variables, and residual variances within the factor model. Modification of the initial factor model due to a less than satisfactory goodness-of-fit index (GFI) (0.848) resulted in the inclusion of four additional relationships between observed variables and a second factor and six paired relations among the residual variances. The modified factor model had a satisfactory GFI of 0.95. These findings support the development of a 15-item self-administered ADL Self-Care Scale. The scale can be used to screen ADL functional levels in MS persons or to monitor changes in their ADL over time.