A new performance index is proposed for managing the service performance of a service system based on the sample data of categorical scale satisfaction on multidimensional service elements. With the new performance index, one can judge whether the level of service performance deviates from the desired one, and recognise the service elements that fail to meet service standards when the level deviates from the desired one. To prevent unavoidable sampling errors that results in incorrect conclusions, a hypothesis-testing procedure of the new performance index is proposed. The contributions and innovation points of this study are: (1) the categorical data used in this study is neither limited to ordinal, nor limited to nominal scale, and thus the approach can provide a wider range of applications; (2) the new method not only monitor the overall performance of a service system but also give an out-of-control interpretation when detecting disappointed service performance.

A new performance index is proposed for managing the service performance of a service system based on the sample data of categorical scale satisfaction on multidimensional service elements. With the new performance index, one can judge whether the level of service performance deviates from the desired one, and recognise the service elements that fail to meet service standards when the level deviates from the desired one. To prevent unavoidable sampling errors that results in incorrect conclusions, a hypothesis-testing procedure of the new performance index is proposed. The contributions and innovation points of this study are: (1) the categorical data used in this study is neither limited to ordinal, nor limited to nominal scale, and thus the approach can provide a wider range of applications; (2) the new method not only monitor the overall performance of a service system but also give an out-of-control interpretation when detecting disappointed service performance.