Owning to the heterogeneity of services, service failures are inevitable to happen in service encounters. Therefore, recovery efforts play important roles in mataining the relationship with customer. Recovery paradox (RP) refers to the situation in which the customer who experiences a failure followed by a superior recovery rates a service as high as or even higher than s/he would rate a service involving no failure. By contrast, Double deviation (DD) refers to the situation in which inappropriate and/or inadequate recovery results in magnification of the negative evaluation. The RP and the DD are essentially symmetrical because both of them represent a phenomenon in which the recovery contributes to customer¡¦s evaluation more heavily than the initial failure dose. Although numerous studies devotes to the question whether the RP and DD exist, very few are trying to answer the question of why they exist. Thus, the main purpose of this research is to conceptually and empirically compare the RP and the DD in order to uncover the potential asymmetry, as well as to understand why recovery influences evaluation more greatly? Because the investigated conditions are not easy to be identified in the real world, a scenario-based quasi-experimental design is chosen. The data is collected from customers actually engaged in the target services. Customers are asked to answer questions about an organization they have recently patronized and then evaluate experimentally-generated scenarios in a restaurant setting to understand whether a negative discrepancy can really magnify the customer¡¦s evaluation toward an identical following event. The result reveals that after a negative-discrepant first event, a positive-discrepant second event is evaluated more positively than non-discrepant second event, but a negative-discrepant second event is evaluated less negatively than a non-discrepant one. That is to say the result supports RP, but doesn¡¦t support DD.

No description available.