AN EMPIRICAL STUDY OF THE QUALITY AND CONTEXT-SPECIFIC APPLICABILITY OF CUSTOMER SATISFACTION MEASURES

Abstract

For the past two decades, there has been a proliferation of research in satisfaction. However, the area of measurement has been largely neglected. There is a need for accurate and precise measurement of customer satisfaction, in order for both academics and practitioners to better manage satisfaction. By investigating into the quality of several satisfaction measures and the context-specific applicability of these under differing product benefits (utilitarian vs. hedonic) and differing levels of consumer involvement with a product (highly vs. lowly involved), this study sought to advance knowledge into this area. A survey design was used for this study. Based on a pre-test, two products (ice cream restaurant and A TM services) were selected as the research contexts for this study. A total of nine overall satisfaction measures, including six single-item and three multi-item, were examined in this study. The study first provided an examination of the quality of the measures in terms of their reliability, validity, skewness coefficients, error variances and satisfaction loadings. Initial findings supported the popular notion that multi-item measures performed better than single-item measures. Next, the measures were examined in terms of the amount of cognitive and/or affective factors they captured. Contrary to popular belief, the results show that all the nine measures captured both factors at varying degrees, with some capturing both factors equally. Finally, the study investigated whether the selection of appropriate satisfaction measures should be a function of product benefits and consumer involvement. Results disproved the first notion but provided support for the second. Good quality measures were shown to be equally applicable for the two products tested, regardless of the combinations of product benefits. On the other hand, the use of more detailed measures was shown to be more appropriate for highly involved subjects than lowly involved subjects.