Validity of Two Self-care Instruments for the Elderly

The aim of this study was to address concurrent and construct validity of two instruments, the Self-care Ability Scale for the Elderly (SASE) and the self-report form of the Appraisal of Self-care Agency Scale (ASA-A), by testing them with home-dwelling, elderly Swedish patients. The two self-report instruments, including some additional new questions, were mailed to an age-stratified random sample of 420 elderly subjects and were finally completed by 125 of them. Bivariate and multivariate statistical methods were used in the analyses. An analysis of the non-respondents revealed no significant differences between them and the respondents. Standardized regression coefficients for reported abilities/activities on the scores were 0.71 for SASE (p < 0.0001) and for ASA-A 0.51 (p < 0.0001). SASE and ASA-A were found to measure related but not identical concepts (r = 0.69). Sensitivity and specificity ratings for managing without any help from official and unofficial caregivers were 68% and 72% respectively for SASE, and 50% and 64% for ASA-A. Individuals of 80+ years of age had lower scores than younger respondents on the SASE (p < 0.001) and ASA-A (p < 0.05) scales. Respondents living independently had higher scores on both scales than those living in institutions (p < 0.001). Lower scores on both scales were found for respondents who were not in good health compared with those who were in good health (p < 0.001). A five-factor solution for SASE explained 70.2% of the variances, and an eight-factor solution explained 67.9% of the variances for ASA-A. The results of this study confirm that the SASE and ASA-A assessments substantially measured what they were intended to measure in the studied group of lucid, elderly subjects.