Submission 182
Learning Norms from Repeated Observations: Behavior of the Majority or Majority of Behaviors?
PS2-G08-02
Presented by: Thomas Karl Alfred Woiczyk
How do individuals infer social norms from observed behaviors within a reference group? Standard models assume that perceived descriptive norms align with the behavior of the majority, yet real-world settings often feature discrepancies between majority behavior and the most frequently observed behavior—what we call "common behavior." We propose that when such discrepancies arise, individuals are more likely to infer norms from common behavior rather than the majority. This divergence can be explained by two key mechanisms: (1) the structure of the information environment, which determines how often individuals encounter specific behaviors, and (2) memory constraints, which lead individuals to recall the most frequent behaviors rather than a strict majority-minority distinction.
To test this, we conducted four pre-registered online experiments (N = 2,104).
Study 1 simulated an arrival-at-work scenario where participants repeatedly observed others arriving in the morning. Despite the majority consistently exhibiting one form of arrival, participants were more likely to misperceive the most common arrival time as the majority's and followed it in their own choices.
Study 2 employed a Dictator Game, where participants saw varying distributions of generous and selfish transfers made by previous players. This study was incentivized, and again, perceived norms aligned more closely with the most frequently observed behavior rather than the majority choice, influencing participants' subsequent decisions. These findings provide robust evidence that individuals’ perceptions of social norms depend more on frequent behaviors than the majority engaging in them.
In Study 3, we tested whether making the majority behavior more salient would improve participants' ability to accurately perceive descriptive norms. Participants, playing the role of employees in a foreign country, observed colleagues' greeting behaviors. We manipulated the presentation mode to highlight the majority behavior. Results showed that when the majority behavior was more apparent, participants were more likely to recall and align their greetings with the majority.
Study 4 extended the investigation within the same greeting form paradigm by manipulating the relevance of majority and minority subgroups based on gender. The results showed that participants were more likely to align with behaviors exhibited by individuals they deemed more socially relevant—specifically, those of the same gender—regardless of whether these individuals were in the majority or minority. This finding supports the idea that norm perception is shaped by social identity and establishes a key boundary condition for our main findings. It demonstrates that when the perceived relevance of a subgroup is strong enough, individuals may prioritize aligning with that subgroup’s behavior, even if it contradicts the behavior of the majority.
Our results contribute to the literature on norm formation and behavioral economics by showing that people do not always follow the majority when inferring social norms. Instead, they rely on the most frequently encountered behaviors, leading to systematic deviations from majority-based norm perceptions. These findings have broad implications for understanding coordination failures, pluralistic ignorance, and the persistence of misinformation in social networks.
To test this, we conducted four pre-registered online experiments (N = 2,104).
Study 1 simulated an arrival-at-work scenario where participants repeatedly observed others arriving in the morning. Despite the majority consistently exhibiting one form of arrival, participants were more likely to misperceive the most common arrival time as the majority's and followed it in their own choices.
Study 2 employed a Dictator Game, where participants saw varying distributions of generous and selfish transfers made by previous players. This study was incentivized, and again, perceived norms aligned more closely with the most frequently observed behavior rather than the majority choice, influencing participants' subsequent decisions. These findings provide robust evidence that individuals’ perceptions of social norms depend more on frequent behaviors than the majority engaging in them.
In Study 3, we tested whether making the majority behavior more salient would improve participants' ability to accurately perceive descriptive norms. Participants, playing the role of employees in a foreign country, observed colleagues' greeting behaviors. We manipulated the presentation mode to highlight the majority behavior. Results showed that when the majority behavior was more apparent, participants were more likely to recall and align their greetings with the majority.
Study 4 extended the investigation within the same greeting form paradigm by manipulating the relevance of majority and minority subgroups based on gender. The results showed that participants were more likely to align with behaviors exhibited by individuals they deemed more socially relevant—specifically, those of the same gender—regardless of whether these individuals were in the majority or minority. This finding supports the idea that norm perception is shaped by social identity and establishes a key boundary condition for our main findings. It demonstrates that when the perceived relevance of a subgroup is strong enough, individuals may prioritize aligning with that subgroup’s behavior, even if it contradicts the behavior of the majority.
Our results contribute to the literature on norm formation and behavioral economics by showing that people do not always follow the majority when inferring social norms. Instead, they rely on the most frequently encountered behaviors, leading to systematic deviations from majority-based norm perceptions. These findings have broad implications for understanding coordination failures, pluralistic ignorance, and the persistence of misinformation in social networks.