Doctoral Dissertations

Date of Award

8-2025

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Retail, Hospitality, and Tourism Management

Major Professor

Youn-Kyung Kim

Committee Members

Sejin Ha, Michelle Childs, Matthew Pittman, Xiaopeng Zhao

Abstract

As artificial intelligence (AI) technologies increasingly shape the retail landscape, AI-powered retail service robots (AI-RSRs) are becoming prominent tools for enhancing customer experiences. However, consumer responses to these technologies are complex and influenced by both psychological perceptions and communication design. This dissertation integrates the Push–Pull–Mooring (PPM) framework and Media Richness Theory (MRT) to examine the antecedents and optimization of AI-RSR acceptance through two empirical studies.

Study 1, a field experiment conducted at a university campus store, tested a structural model grounded in PPM theory. Results revealed that perceived usefulness and coolness significantly enhanced consumers’ attitudes toward AI-RSRs, while eeriness exerted a significant negative influence. Additionally, perceived outdatedness—reflecting dissatisfaction with traditional retail environments—was a significant positive predictor. In contrast, social anxiety and anxiety toward AI-RSRs were not significant factors.

In addition to psychological mechanisms, Study 1 also examined the effects of robot communication richness on consumer perceptions. Drawing on MRT, three dimensions—multiple cues, language variety, and personal focus—were experimentally manipulated. Among the dimensions, language variety showed a statistically significant main effect on perceived usefulness, with the expansive (i.e., overly expressive and humorous) speech style yielding lower usefulness ratings than more moderate styles. While the effects on coolness, social presence, and eeriness were not statistically significant, the default communication style—moderately expressive and without name personalization—consistently produced the most favorable mean scores across all perception dimensions. Notably, eeriness was lowest in the default condition, suggesting a potential threshold where excessive richness may begin to hinder consumer comfort.

Study 2 replicated the PPM-based structural model in a broader digital context using an online survey. Results closely mirrored those of Study 1, reinforcing the robustness and generalizability of the proposed model.

Together, Study 1 and Study 2 offer theoretical insights into the psychological and communicative mechanisms that shape consumer acceptance of AI-RSRs. The findings contribute to the literature on human–AI interaction by identifying critical antecedents and demonstrating the nuanced role of communication richness. Practically, this research provides actionable guidance for designing more effective, human-centered AI-RSRs that are not only functional and engaging but also socially acceptable and psychologically comfortable.

Download

Available for download on Friday, August 15, 2031

Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS