Previous research has analysed service quality mainly with single- and two-predictor models. The single-predictor models indicate the disconfirmation-only model and the performance-only model, whereas the two-predictor models employ a pair of predictors in the set of three predictors (expectation, performance, and disconfirmation). This paper proposes a general three-predictor model nesting the alternative single- and two-predictor models and then shows linkages between the alternative models. However, previous research has used and interpreted the alternative models without considering the linkages between the alternative models. This paper shows that this interpretation may lead researchers to incorrect conclusions on a consumer’s evaluation of service quality because different statistical implications may result from the linkages. Accordingly, this paper ends with the recommendation that one should measure service quality with the two-predictor models and interpret the effects of predictors based on the linkages between the theoretical three-predictor model and the two-predictor models.