M.A. Thesis / In progress
AI News Anchors’ Smiles, Bowing Depth, and News Credibility: A South Korea–China Comparison of Indirect Effects Through Perceived Eeriness
AI news anchors are already on air in both South Korea and China, but audiences may not read their nonverbal behavior the same way. This thesis asks whether two culturally loaded greetings — a smile and a bow — make an AI anchor more credible, or push it into the uncanny valley. Two between-subjects experiments vary smile intensity (Study 1) and bowing depth (Study 2) for Korean and Chinese participants. Perceived eeriness is tested as the pathway from each cue to news credibility, with country as the moderator.
- Research question
- Do an AI anchor’s smile and bow raise news credibility, or lower it by making the anchor feel eerie — and does the answer differ between Korea and China?
- Theory
- Computers Are Social Actors (CASA) paradigm · expectancy violation theory · the uncanny valley
- Design
- Two 3 × 2 between-subjects experiments (three cue levels × two countries)
- Measures
- Perceived eeriness (Ho & MacDorman) · message credibility (Appelman & Sundar)
- Analysis
- Moderated mediation (PROCESS Model 8)
- Status
- Proposal approved · experiments in preparation
The model is specified separately for Study 1 and Study 2.
Paths a and b form the indirect pathway through perceived eeriness; c′ is the direct effect on credibility. Country moderates a and c′. Dashed arrows show moderation.
