Effectively communicating uncertainty: The persuasive impact of different types of hedges
研究通过实验和机器学习分析数百万条在线评论，发现模糊限制语（如“可能”“大概”）的视角（个人化vs一般化）和所暗示的可能性高低通过影响感知信心来改变说服效果，并指出品牌可通过拟人化代理利用这一效应。
Abstract Communicators often hedge. Consumers say a restaurant might have good mimosas or that a product is probably the best while Amazon suggests movies they think you'll like. But might these different hedges have different effects on persuasion? We suggest that (1) whether a hedge takes a personal (vs. general) perspective and (2) the likelihood suggested by a hedge both play important roles in determining hedging's persuasive impact. A multi‐method investigation combining experiments with machine learning of millions of online reviews supports these hypotheses. Together, they demonstrate that the effects of both hedge perspective and hedge likelihood are driven by a common mechanism: perceived confidence. Hedges that involve personal perspective or higher likelihood increase persuasion because they suggest communicators are more confident. Moreover, we demonstrate that hedging can protect communicators from backlash without sacrificing persuasive impact and highlight that brands can leverage hedging's persuasive impact when speaking through anthropomorphized agents. This work contributes to the literature on language in marketing, showcases how subtle linguistic features impact perceived confidence, and has clear implications for anyone trying to be more persuasive.