Electronic Theses and Dissertations

Della Clark

Date

2024

Document Type

Dissertation

Degree Name

Doctor of Philosophy

Department

Business Administration

Committee Chair

Sandra Richardson

Committee Member

Abhijit Guha

Committee Member

Gregory Boller

Committee Member

Roger Kreuz

Abstract

In this dissertation, I examine consumer expectations of service agents in different contexts across two essays. This is the first empirical research to examine differences in the probabilistic expectations of an event (within a service recovery context where outcomes are specific, known, and desired by the consumer) is known occurring when dealing with different agent types (human vs. artificial intelligence (AI)). While the proliferation of ‘A New Age of AI’ news narrative is warranted and certainly the reality of service contexts, I aim to show that consumers systematically perceive events as more likely to occur when dealing with human agents, as opposed to their AI counterparts. In the first essay, I examine an underlying bias toward human (vs. AI) interactions in service interactions. First, I demonstrate that agent type (human vs. AI) affects customers’ probabilistic expectations of an event occurring across four studies, including an incentive-compatible study design. Then, I examine the mechanism underlying this effect, and further, provide mediation process support via moderation. The findings suggest that consumers’ probability judgments may be systematically biased based on agent type due to differing perceptions of psychological distance. Within the second essay, I plan to examine different agent types (human vs. AI) in managerial and hiring contexts. Highlighting the relevance of AI in hiring and managerial contexts is evidenced by new and ongoing regulatory efforts, i.e., 820 ILCS 42/ Artificial Intelligence Video Interview Act. As AI expands further into the human resources (HR) realm, I plan to answer the question of whether interactions with human vs. AI agents sustain this human bias for probabilistic expectations. Expressly stated, do interactions of employees and potential hires with different agent types (human vs. AI) affect their expectations? These two essays contribute to the growing literature of the effects AI expansion has on consumer interactions as well as to the broader literature on service recoveries and ethics in hiring. Uncovered within Essay I, I demonstrate the effect of a ‘human heuristic,’ in which, consumers have higher probabilistic expectations of events occurring across a variety of service contexts. Rather than suggest that all AI should be replaced by humans or vice versa, this research supports a growing body of work that suggests hybrid agent models, while acknowledging the ‘human heuristic’ bias.

Comments

Data is provided by the student.

Library Comment

Dissertation or thesis originally submitted to ProQuest.

Notes

Embargoed until 5/1/2026

Download

Available for download on Friday, May 01, 2026

Share

COinS