JMIR Publications

ABSTRACT

Background:

Incidents of online public shaming can have devastating consequences for those who are shamed, but how those who witness shaming behaviors react is unclear. When considering online public shaming, it is crucial to be aware of the context in which it occurs. Implicit norms which govern these contexts and evoked emotions can influence what is deemed as acceptable behavior when witnessing public shaming. However, prior work has not examined the role that perceived social acceptability of the shaming content and emotional arousal may have in explaining social media behavior after witnessing online shaming incidents.

Objective:

We pose three research questions to explore and characterize people’s reactions to witnessing online public shaming: 1) does perceived social acceptability predict subsequent behavior to tweets; and 2) are participants’ emotional reactions to the tweet associated with self-assessed likelihood to engage in certain behaviors; and 3) what do participants’ explanations of their reactions to the tweets illustrate about their views about appropriate online behavior?

Methods:

We conducted a between-subjects experimental design in which participants saw one of four tweets: shaming of a couple holding a wedding that became a super spreader event (Shaming Condition) and three control tweets: one expressing positive sentiment about a wedding during COVID (Wedding Condition), a public service announcement providing information about obtaining COVID-related information (COVID Condition); and a tweet expressing neutral sentiment (Control Condition). To answer the first two questions, we constructed regression models with replying/commenting, sharing, and liking as the dependent variables, and factors including demographic characteristics, social media behaviors, tweet type, perception of social acceptability, and emotions as the independent variables. To characterize how participants interpreted the online act, we performed an inductive qualitative analysis of open-ended responses explaining their reactions to the tweet they saw, and then compared the prominence of the themes across the conditions using analyses of variance (ANOVAs).

Results:

We invited 800 participants, and 742 participants completed the entire study. With respect to the first research question, the perception of a tweet as socially acceptable increased all forms of social engagement, with increased liking, sharing, and commenting. With respect to emotions, positive emotions were associated with greater tendency to engage in all three behaviors, whereas fear, shame, and anger involved more interactions between the tweet type, emotion, and the subsequent behavior (liking, sharing, or commenting). Last, qualitative analysis of the open-ended responses yielded six factors affecting subsequent behavior: Environment, Sense-making and Assessment, Message, Tone, and Other, with significant differences in the themes across the tweet types.

Conclusions:

This study augments our understanding of people’s responses to online public shaming, including diverse factors to interpret and make decisions concerning their own subsequent actions. Considering these factors in platform design can potentially mitigate negative consequences.