The factor structure of customer satisfaction

There is growing evidence that service quality attributes fall into three categories of factors that have a different impact on the formation of customer satisfaction. However, it is not clear which analytical procedure best identifies these factors. Vavra proposed a two‐dimensional importance grid based on customers’ self‐stated importance and derived importance using regression analysis. It is based on the assumption that there is a difference between self‐stated and derived importance and that by combining these importance weights, three groups of product or service attributes can be identified. Using data collected to measure customer satisfaction with the service of the IT department of a hospital, the authors test the underlying assumptions of the importance grid. They seem to be correct. When the results are compared with the penalty‐reward contrast analysis developed by Brandt, the two methods do not yield the same results. Therefore, the convergent validity of the importance grid has to be questioned. The paper closes with a discussion of the implications for research and practice.