Using Chernoff faces to portray service quality data

Abstract

This paper presents an attempt to demonstrate the use of faces as a statistical data presentation technique in the portrayal of service quality data. The statistician Chernoff originally introduced a technique for the representation of multivariate data by means of the human face. It is helpful in that widely divergent facial features are shown, each of which can be associated with a different variable.

This study uses data obtained from an experimental research design, in an attempt to compare the facial presentation of multivariate service quality data. The data set was gathered across two groupings of customers which in turn was subdivided into users and non users of the service which served as input for the preparation of the Chernoff faces. The results offer a more concise and holistic interpretation of the service quality picture. It is almost certain that any individual would be able to identify the most pleasing face relative to the others.