An Empirical Basis for Domains in the Analysis of Dependency in the Activities of Daily Living (ADL): Results of a Confirmatory Factor Analysis of the Cleveland Scale for Activities of Daily Living (CSADL)

Abstract

This study examines the empirical basis for dividing activities of daily living (ADLs) into broad behavioral domains. Following an exploratory factor analysis of results from the Cleveland Scale for Activities of Daily Living (CSADL), a two-factor model of ADLs was tested in a confirmatory factor analysis using two groups of demented patients. The model, which included a Basic and an Instrumental factor, was confirmed. Comparison of the item composition of the CSADL factors with a priori domains used by other ADL scales suggested that the results of the present study are broadly generalizable. The two factors are correlated, and cautions were raised concerning the likelihood that many ADL items may reflect both basic and instrumental domains.

ACKNOWLEDGMENTS

This study was supported in part by grant AG08012 from the National Institute of Aging, a grant from the Eli Lilly Company, grant NH4344–01 from the National Institute of Mental Health, and an Alzheimer Center grant from the Ohio Department of Aging. We acknowledge the help of the staff of the University Memory and Aging Center in gathering and providing the data for the analysis in the present study.

Notes

¹Scale items are published in Patterson and Mack (2001). The scale is available from Health and Psychosocial Instruments (HaPI), E-mail: [email protected], as Cleveland Scale for Activities of Daily Living, Accession Number 197895.

²Bathing, Toileting, Personal Hygiene, Dressing, Eating (including Meal Preparation), Mobility, Medications, Shopping, Travel, Hobbies/Employment, Housework/Home Maintenance, Telephone, Money Management, Communication Skills, Social Behavior.

^a Loadings < .10 are not presented.

³“Error variance” in this context includes (a) actual variance attributable to error of measurement, (b) variance specific to a given item, and (c) variance shared by a two specific items distinct from the variance they share that is common to the factor. It is this third source of variance that justifies freeing these parameters.