Responding to Causal Uncertainty in the Twitterverse: When Abstract Language and Social Prominence Increase Message Engagement

Events that capture the public’s attention are saturated with causal uncertainty (i.e., uncertainty about why things happen). Yet, little is known about the type of language that is favored when communicating such events. We study this issue in the context of social media, as social media platforms are powerful tools for communication in politics and business, especially during times of heightened causal uncertainty. When causal uncertainty is high, people should evaluate more favorably messages that help relieve such uncertainty. Germane to our research, previous work has shown that abstract thinking helps reduce causal uncertainty, and that the goal to reduce causal uncertainty increases the desire for abstract thinking. We extend these findings to better understand the role of causal uncertainty in social interactions and message engagement. Using data from natural social media communications and experiments based on high-profile events, we demonstrate that individuals react more positively to abstract (vs. concrete) messages in situations of heightened causal uncertainty. Importantly, this effect is stronger when the message source is socially prominent, and hence, the message carries greater diagnosticity. Our results indicate that increasing the abstractness of a tweet by a standard deviation increases the average number of likes it receives by 10.25% and retweets by 4.07%. When originating from a source with a standard deviation of more followers than the average user, these messages receive 15.80% and 5.98% more likes and retweets, respectively. Beyond the clear managerial implications of our findings, we provide unique insights into why simplified causal explanations for complex problems may win over an audience during times of causal uncertainty.