UC San Diego

This dissertation contains three papers that together shed light on how and why the consideration of negative stimuli (i.e., information, outcomes, language) shapes individuals’ preferences and behavior.

Chapter 1 investigates how the consideration of negative information about a scholar—unrelated to their work—affects other scholars’ citing behavior. Academics use citations to acknowledge the contribution of past work and promote scientific advancement. However, analyzing citation data of 32,025 publications spanning 18 academic fields, we find evidence suggesting citations may also serve as a currency to reward and punish scientists’ morality. We find that, relative to controls, the citation rates of scholars accused of sexual misconduct decrease after the accusations become public. Interestingly, this citation penalty is larger than the one incurred by scientists accused of scientific fraud. Our findings suggest that, in addition to serving the purpose of maintaining intellectual integrity and promoting scientific advancement, citation decisions are also driven by scholars’ attitudes toward the publications’ authors.

Chapter 2 demonstrates how and why the consideration of negative outcomes impacts people’s attributional preferences. There are two streams of literature that address attributional preferences: self-determination and self-serving preferences. While these two theories make the same prediction for individuals’ attributional preferences over positive outcomes, they make competing predictions for attributional preferences over negative outcomes. Self-determination maintains that people prefer to have agency over negative outcomes. Self-serving preferences, in contrast, stipulate that people prefer to concede agency over negative outcomes. In eight preregistered experiments (N = 3,946), we reconcile these seemingly inconsistent attributional preferences over negative outcomes. First, we test these competing predictions and find that—consistent with self-determination—people would rather “own” their negative outcomes than externally attribute them. Overplacement (people’s belief that they perform better than others) and the impact bias (the belief that “owned” negative outcomes hurt less than when they are caused by oneself) cannot explain this preference. Instead, we find that reducing the saliency of agency moderates the preference for agency over negative outcomes. More interestingly, we find that sharing agency reverses attributional preferences: while people prefer assuming agency over negative outcomes when these are exclusively caused by a sole agent (either themselves or somebody else), they prefer attributing agency to others when negative outcomes are jointly caused by multiple agents (both themselves and somebody else).

Finally, chapter 3 examines how different cognitive processes elicited by negative language affect a message’s efficacy. Across five preregistered field and lab experiments (N = 22,024), we demonstrate when and how using an easily reversible (i.e., bi-polar) word in a statement, rather than a non-reversible one with the same meaning, engages different cognitive processes and leads to different outcomes. In particular, when a statement containing a bi-polar word is processed as a negation (i.e., opposing a claim rather than affirming it), a slower more elaborate cognitive process occurs. We show that this results in lower judgment confidence, and a lower likelihood to act on the message. In addition, we find that this more elaborative process also leads to weaker attitudes towards the message source. Our findings advance consumer theories by shedding light on the ways in which linguistic elements of communication impact judgments and real-world behaviors. They additionally offer practical persuasive messaging strategies for those engaged in a range of marketing and policy communications.