Date of Award
Doctor of Philosophy (PhD)
Dr. Xiaojing Sheng
Dr. Reto Felix
Dr. Michael Minor
This dissertation examines the effects of viewing and hearing service failure and recovery interactions between self-service technologies (SSTs) and customers on third-party observers. Third-party observers are those that observe the actions of other customers as they interact with a firm in some way. These observations could occur in the present (face-to-face) or past (watching a video or listening to an audio recording). Third-party observers provide a unique perspective as they can provide a more neutral perspective of a failed service transaction and recovery as compared to the involved customer and firm.
This dissertation contributes to the SST literature and to theory by creating a virtual service agent acceptance model. An important grounding theory used is attribution theory. Simply, attribution theory (Heider 1958; Weiner 1985, 2006) posits that we are motivated to attribute meaningful causes to action and behavior. The theory suggests that when faced with the task of appraising an outcome, we broadly make either dispositional attributions, which reside firmly within the individual, or situational attributions, which refer to external factors outside of the individual. Attribution theory has been extended to the service context to explain where failures are attributed, and the causal inferences made by customers when a service failure occurs. While much literature has been written on the attribution process with the interaction between human employees and customers, the literature remains sparse on attribution theory as it relates to third-party observers viewing interactions between SSTs and customers. Additionally, much of the previous research has assumed the locus of causality to be unambiguous, assuming the cause of service failures lies within the firm or employee. This may not always be the case with new SSTs, such as avatars, and the cause of the failure may indeed lie within the customer or with situational aspects. I intend to fill this gap with the dissertation and delve into the mechanisms that drive attribution. As artificial intelligence increases the capabilities of self-service technology and these SSTs progress to be closer and closer to being indistinguishable from real humans, this research is important to managers of firms as they will make decisions on employing self-service technologies, human employees or a mixture of both working side by side.
Study 1 specifically explores a service failure and recovery situation between a customer and an SST (i.e., an avatar). Study 1 finds main effects of customer attributes, avatar attributes and important interaction effects of service elements that impact attributions of failure. Further, Study 1 shows how these attributions of failure impact the third-party observer’s satisfaction toward the avatar. Furthermore, Study 1 shows how their perceived satisfaction affects their future approach intention of avatars. These are important findings as they show that third-party observers are able to evaluate service interactions that fail and subsequent recoveries to the failure and use this information to determine their perceived satisfaction and potentially whether or not they intend to use the services they observed in the future.
Lastly, Study 1 adds to the attribution theory literature as it left the locus of causality dimension ambiguous allowing the respondent, acting as the third-party observer, to attribute the failure to not only the firm or avatar, but also the customer. This is an important gap to fill as it allows the observer the opportunity to attribute failure to the customer where much of the previous literature has not.
Study 2 further extends the virtual service agent acceptance model as it uses appraisal theory to ground the prediction concerning negative emotions and coping responses to an observed failure. Simply, appraisal theory predicts that when observing an event, an observer appraises the event, experiences emotions through this appraisal and copes with the observed event (Lazarus, 1991; Lazarus and Smith, 1988).
Study 2 specifically shows the effects of service elements on emotions and coping resources. Moreover, it shows the effects of these coping resources on choice behavior toward SSTs of third-party observers. Study 2 adds to and extends Study 1 and adds robustness to the findings.
Together, these studies contribute to the literature in three ways 1) Adding to the emerging AI powered SST literature, 2) blending the established theories of attribution and appraisal theory to form the virtual service agent acceptance model to explain and predict the acceptance of SSTs by third-party observers, and 3) adding to the "others" literature showing the importance of the effects of observing a customer interact with the firm on third parties in a service interaction.
Murray, Ross Paul, "Self-Service Technologies that Appear Human Interacting with Customers: Effects on Third-Party Observers" (2022). Theses and Dissertations - UTRGV. 1075.