Abstract
A study into the attitudes and evaluations of 195 older adults between the ages of 65 and 95 years attending a hospital-based low-vision clinic has been carried out, with special emphasis on patients' satisfaction judgements. Measures comprised aspects of visual function, clinical status, and a battery of quality-of-life questionnaires including the SF-36 social functioning scale, the Life Satisfaction Index, the Nottingham Adjustment Scale for information on psychological characteristics, a measure of Intrinsic Religious Motivation, and information on social support. Four different measures of satisfaction were used: a person's satisfaction with their low-vision aids, satisfaction with reading ability, satisfaction with the low-vision service, and satisfaction in general. A fifth 'overall' measure of satisfaction was derived as the sum of responses on these four scales. The data were analysed in a number of ways, including principal component analysis, logistic regression, and a classification and regression binary tree-growing algorithm. The principal predictors of overall satisfaction were reading accuracy, the degree of psychological adjustment to vision loss, and an aspect of physical functioning. These were similar by both regression methods, but the classification and regression binary tree-growing algorithm demonstrated that the predictors of dissatisfaction were not the same as the predictors of satisfaction. The results show that satisfaction scales are complex and are not univariate and the uniqueness associated with different satisfaction scales is more apparent than is their shared meaning. The findings emphasise the importance of taking great care in selecting scales of satisfaction for evaluating service outcome.