RESUMO
Chatbots, a type of virtual AI entity designed to emulate human conversation, are gaining prominence in business and consumer domains. This research aims to consolidate extant literature focusing on a pivotal aspect: the human-likeness of chatbots. Employing three fundamental themes as organizational pillars - chatbot as a non-human entity, chatbot as a human-like entity, and chatbot as an ambiguous agent - we aim to not only spotlight important findings but also chart out potential trajectories for future exploration. By delving into the intricacies of chatbot-consumer interaction, we seek to shed light on unexplored dimensions of marketing research, ultimately enhancing our understanding of this evolving field.
RESUMO
While the global pandemic highlighted the importance of adhering to boundaries (e.g., social distancing rules), compliance with these boundary-imposing measures has been politically divided. This research proposes one reason that may underlie the observed ideological asymmetries toward COVID-19 prevention measures and boundaries in general: Conservatives and liberals may fundamentally differ in how they construe boundaries. Supporting this prediction, Studies 1a-1d and two follow-up studies (n = 3,231; Studies 1a-1c and follow-up studies: Amazon Mechanical Turk and Prolific users, Study 1d: U.S. students) demonstrate that identifying with political conservatism (vs. liberalism) increases the likelihood to construe boundaries as restrictions. We further show that, due to conservatives' greater preference for order, structure-related words carry a more positive connotation among conservatives versus liberals (Study 2: n = 744; MTurk users). Capitalizing on this finding, we demonstrate that linguistic framing that highlights the structure-providing function of a boundary (e.g., a social distancing sign can "structure" customer flow in a restaurant) can reduce the salience of its usual restrictive aspect and hence effectively improve conservatives' attitudes toward the boundaries (Study 3: n = 740; MTurk users). (PsycInfo Database Record (c) 2023 APA, all rights reserved).