Degree Name
Master of Science
Department or Program
Mass Communication and Media Arts
Advisor
Ryoo, Yuhosua
Abstract
Will new media techniques, such as artificial intelligence (AI), help refresh public safety advertising campaigns and help better target specific populations, and aid in persuasive, preventative marketing? This paper used hypocrisy induction as a persuasive tool for standalone artificial intelligence chatbots to test potential behavioral change in the context of marijuana. This research further tested whether the chatbots' gender and language styles impact how persuasive and effective the chat agents are perceived to be using hypocrisy induction. An online experiment conducted with 705 participants (Mage = 42.9, 392 women). where participants interact with a chatbot that is manipulated as male/female and uses formal/causal language. Half of the participants received the hypocrisy induction manipulation. hypocrisy induction is more effective when chatbot gender and linguistic styles are appropriately paired. Participants in the hypocrisy induction condition exhibited higher WTP than those in the non-hypocrisy induction condition when the chatbot they interacted with was female in gender and used casual language. However, hypocrisy induction increased WTP than those who did not receive the hypocrisy induction manipulation when the gender of the chatbot they interacted with was male and used formal language. To the researchers' knowledge, this is among the first studies testing the persuasive power of hypocrisy induction using new media platforms in public safety and health advertising in marijuana studies. Findings not only help to shed light on the persuasiveness of gender and language in standalone chatbots but also provide practical implications for practitioners on the future usage of chatbots.