Free
Brief Report  |   January 2004
Convergent and Predictive Validity of Three Scales Related to Falls in the Elderly
Author Affiliations
  • Anita Hotchkiss, MS, OTR/L, is Instructor and Fieldwork Coordinator, Gannon University, 109 University Square, Erie, Pennsylvania 16541; hotchkis001@gannon.edu
  • Andrea Fisher, MS, OTR/L, is Staff Therapist, Miami Valley Special Education Regional Resource Center, Dayton, Ohio
  • Randi Robertson, MS, OTR/L, is Staff Therapist, Hamot Medical Center, Erie, Pennsylvania
  • Amy Ruttencutter, MS, OTR/L, is Staff Therapist, Norwood City Schools, Cincinnati, Ohio
  • Julie Schuffert, MS, OTR/L, is Staff Therapist, MedCentral Health System, Mansfield, Ohio
  • David Bruce Barker, ScD, is Director, Sociology Program, Gannon University, Erie, Pennsylvania
Article Information
Geriatrics/Productive Aging / Departments / Brief Report
Brief Report   |   January 2004
Convergent and Predictive Validity of Three Scales Related to Falls in the Elderly
American Journal of Occupational Therapy, January/February 2004, Vol. 58, 100-103. doi:10.5014/ajot.58.1.100
American Journal of Occupational Therapy, January/February 2004, Vol. 58, 100-103. doi:10.5014/ajot.58.1.100
Abstract

The purpose of this study was to determine if the Activities-specific Balance Confidence (ABC) Scale, Falls Efficacy Scale (FES), and Survey of Activities and Fear of Falling in the Elderly (SAFE) assessments measured fear of falling in the same manner (convergent validity) and to determine if they predicted those individuals who, based upon a previous history of falls, limitation of activity, and not leaving home, had an increased potential for falling (predictive validity). One hundred and eighteen individuals, 60 years of age and older, completed each of the assessments. They self-reported activity restriction, fall history, and the number of times they left the home each week. The convergent validity of each assessment was established by correlating each assessment tool with each of the others. Findings indicated the ABC and FES were highly correlated with each other, indicating they measured similar constructs, and both were moderately correlated with the SAFE, suggesting these assessments measured different constructs. The predictive validity of each instrument in relation to the frequency of falls, limitation of activity, and frequency of leaving the home revealed no individual tool could accurately predict any of these characteristics of the sample. As a result, no one test by itself was able to identify individuals who may be at risk and a candidate for an intervention program.