채용 면접에서 AI와 인간 면접관의 효과 비교: 정보원 신뢰의 매개효과와 과업 유형의 조절효과를 중심으로

초록

This study experimentally examined the differential effects of AI and human interviewers on applicants' negative word-of-mouth intentions, focusing on the mediating effects of source credibility (expertise and trustworthiness) and the moderating effect of task type (fact-based vs. evaluation-based). A 2×2 between-subjects experiment was conducted with 141 university students. The results showed that human interviewers significantly reduced negative word-of-mouth intentions through the mediation of perceived trustworthiness compared to AI interviewers, whereas the indirect effect through expertise was not significant. Next, analysis of the moderating effect of task type revealed that the impact of interviewer type on perceived expertise varied significantly depending on task type, with human interviewers' expertise being more pronounced in evaluation-based tasks. The moderated mediation analysis also showed that the effect of human interviewers in reducing negative word-of-mouth intentions through expertise was stronger in evaluation-based task conditions than in fact-based task conditions. These findings demonstrate that the two dimensions of source credibility play different roles in the relationship between interviewer type and negative word-of-mouth intentions. Trustworthiness exhibited a consistent mediating effect regardless of task context, whereas expertise operated as a moderated mediation pathway that varied depending on task type. This study contributes to theoretical expansion by integrating source credibility theory and task-technology fit theory in the context of AI recruitment, and provides practical implications for organizations considering the adoption of AI interview systems.

키워드