Mittal, Ross and Baldasare (1998) use the previous argument when analyzing the asymmetric impact of attribute performance on overall satisfaction and the intention to repurchase. Mittal et al (1998) use Oliver’s study (1993) as a base. Oliver (1993) in turn builds on the previous models of Westbrook (1987) and Westbrook and Oliver (1991), which explore the emotional source of satisfaction. To evaluate satisfaction with the attributes of a product or service, Oliver (1993) separately asked the consumers about their levels of satisfaction and dissatisfaction on two six-point scales ranging from “not at all” to “very much”. The consumers were invited to indicate their degrees of satisfaction and dissatisfaction with the same attribute. The results of the study indicate that both satisfaction and dissatisfaction have a significant effect on overall satisfaction with a product or service. The variables of satisfaction/dissatisfaction used in the study by Mittal et al. (1998) were obtained based on open telephone interviews in which the consumers were asked about their experiences regarding different features of the service analyzed. The answers were subsequently coded, making for different service feature categories. The categories were then classified as positive or negative. In the case of some features, both a positive and a negative version of the same event or feature were generated. For example, “the doctor listens to the patients” vs. “the doctor is not interested in /does not listen to his/her patients”. In other cases, the positive and negative categories did not have the same reference point. For example, “very friendly with the children” or “does not follow up /does not explain the results”.