Too Many Likes, Not Too Few: Online Popularity Shifts Moral Choices

Being ignored online may leave your moral compass untouched, but a flood of likes could quietly tilt it. That is the surprising conclusion of a new experiment suggesting that the modern currency of social media attention—likes, visibility, and affirmation—may shape how willing people are to endorse harming one person to save others. The study, published in Discover Psychology, found that participants who received an unusually high number of likes on a mock social media platform became more accepting of sacrificial harm in moral dilemmas, while those who were socially excluded decided no differently from those who received a typical amount of attention.

The research, conducted by Mahsa Hazrati of Shahid Beheshti University and Khatereh Borhani of Jena University Hospital, set out to test a question that has been surprisingly underexplored: does being placed into a lower or higher social position causally change the moral decisions people make afterward? Decades of work have shown that social exclusion threatens four fundamental psychological needs—belonging, self-esteem, control, and meaningful existence—and can impair executive functions such as attention, working memory, and self-regulation. Because moral decision-making depends on many of these same cognitive resources, the researchers reasoned that exclusion might plausibly shift moral choices. Yet most prior studies had examined how people morally evaluate exclusion itself, or how exclusion affects ethical behavior, rather than whether the experience of being excluded changes subsequent decisions in classic moral dilemmas.

To manipulate social standing, the team used the online ostracism paradigm, a social media-based alternative to the classic Cyberball ball-tossing game. Participants created a profile with a username, avatar, and a short self-description, and then spent three minutes on a simulated social media website alongside what they believed were eleven other participants. The critical manipulation was the number of likes each participant received on their post: one like for the exclusion condition, five for typical inclusion, and nine for overinclusion. Because the profiles were personalized and written by the participants themselves, the paradigm is thought to induce a stronger sense of ostracism than Cyberball, while better mirroring the digital dynamics of everyday online interaction.

Ninety-one Iranian adults, mostly undergraduates with a mean age of about 24, were randomly assigned to one of the three conditions. After the social media task, they completed a moral dilemmas task consisting of ten scenarios in which the participant was the protagonist and also stood to benefit from the harmful action—so-called self-benefit dilemmas. Five dilemmas involved personal force, such as physically pushing someone, and five involved impersonal means, such as redirecting a threat. For each dilemma, participants first rated how aroused they felt, then made a binary yes-or-no decision about whether to perform the harmful action, and finally rated their willingness on a seven-point scale. The study also measured mood, need threat, rejection sensitivity, and empathy traits, and controlled for age and gender in all primary models.

The manipulation checks confirmed that the social media task worked as intended. Excluded participants reported significantly lower satisfaction of belonging, meaningful existence, control, and self-esteem than both included and over-included participants, and they also reported less positive and more negative affect than the over-included group. Notably, over-included participants reported higher control and self-esteem than typically included participants, suggesting that unusually high affirmation confers a sense of power and status beyond ordinary social connection. These findings align with the idea that overinclusion—receiving disproportionately high positive feedback relative to the group norm—is a distinct and emotionally intense experience, not merely a stronger version of inclusion.

When it came to the moral dilemmas themselves, the results defied the classic theoretical prediction. Dual-process accounts of moral judgment hold that sacrificial choices—endorsing harm to maximize outcomes—depend on controlled cognitive processes that can override emotional aversion to harming others. If exclusion depletes cognitive resources and induces a defensive state of cognitive deconstruction, excluded individuals should have become less willing to accept sacrificial harm. Instead, mixed-effects models showed that the exclusion group made decisions statistically indistinguishable from the inclusion group, in both binary and Likert responses, with Bayesian analyses supporting the null. The overinclusion group, by contrast, showed significantly greater acceptance of sacrificial harm than both other groups, and this pattern held across personal and impersonal dilemmas alike.

The key to the overinclusion effect appears to lie in emotional reactivity. Over-included participants reported significantly lower arousal when confronting the dilemmas, and mediation analyses indicated that this dampened arousal partially explained their greater willingness to endorse harm. According to the affective-inhibition perspective, arousal acts as an emotional brake on harmful action: when people feel the distress of a victim more intensely, they are less inclined to pull the lever, so to speak. Lower arousal therefore weakens harm aversion. Importantly, the authors caution that this pattern reflects reduced harm aversion rather than heightened outcome-maximization—over-included participants were not calculating consequences more carefully, but feeling the weight of harm less acutely. When trait empathy measures were added as covariates, the indirect effect through arousal was attenuated, suggesting that the single arousal rating captures a mixture of situational and trait-level affective tendencies.

Empathy traits themselves told a consistent story. Higher affective empathy—combining empathic concern and personal distress—predicted lower acceptance of sacrificial harm overall, replicating a well-established finding that people who feel others’ suffering more strongly are more reluctant to inflict it. Within the overinclusion group, empathic concern and affective empathy were particularly strong negative predictors of harm acceptance. The authors interpret the overinclusion effect through converging factors: elevated positive mood, a boosted sense of control and power, and status-based self-perception, all of which have been independently linked to greater willingness to accept sacrificial harm. The heightened arousal among excluded participants, meanwhile, fits evidence that short-term ostracism produces hypersensitivity rather than the emotional numbness associated with longer-term rejection.

The study has clear limitations that the authors acknowledge. Only self-benefit dilemmas were used, meaning that rejecting the harmful action would have cost the decision-makers their own lives, so self-preservation motives may have shaped responses. The sample came from a single non-Western cultural context, the arousal measure could not distinguish compassion from anxiety or threat, and no emotionally neutral decision task was included to determine whether the exclusion null effect is specific to moral content or reflects a broader resilience of decision-making. The authors also note that repeated presentation of hypothetical dilemmas may weaken the affective processes engaged by real moral choices, and that the two-stage response format could invite anchoring effects, though consistency checks suggested this did not differ across conditions.

The implications reach beyond the laboratory. On digital platforms, some users—most visibly influencers and content creators—receive disproportionately high approval, visibility, and reinforcement on a daily basis. The authors are careful not to claim that highly visible users are more likely to accept sacrificial harm, since brief laboratory overinclusion may not capture chronic online status. But the findings provide a testable basis for asking whether sustained exposure to unusually high online affirmation is associated with blunted affective responses to harm cues, perhaps through power-related appraisals. For the excluded, the message is almost reassuring: a brief episode of online ostracism, however painful to belonging and self-esteem, did not measurably distort moral judgment in this paradigm. Whether that resilience holds over longer timescales, in other cultures, and in decisions where the decision-maker gains nothing personally remains an open question—one that the authors argue should now be pursued with process-level models that separately estimate consequence sensitivity, norm adherence, and inaction tendencies.

Subject of Research: The effect of online social exclusion and overinclusion on moral decision-making

Article Title: Overinclusion but not social exclusion shifts sacrificial harm acceptance in online ostracism paradigm

Article References: Hazrati, M., & Borhani, K. (2026). Overinclusion but not social exclusion shifts sacrificial harm acceptance in online ostracism paradigm. Discover Psychology, 6(1), Article 252. https://doi.org/10.1007/s44202-026-00881-7

Image Credits: AI Generated

DOI: 10.1007/s44202-026-00881-7

Keywords: social exclusion, overinclusion, ostracism, moral decision-making, sacrificial harm, social media, arousal, empathy, dual-process theory, cyberostracism, moral psychology, likes