Most people think they would defy authority if orders went against their morals. (Josie Elias/Shutterstock)
In a nutshell
- Most people believe they would disobey harmful orders from an authority figure, but this study shows they significantly underestimate their own likelihood of compliance, even when told about the high obedience rates in Milgram’s original experiment.
- The “better-than-average effect” was clearly demonstrated: participants predicted they’d resist more than others would, reflecting a widespread tendency to view ourselves as more moral or courageous than the average person.
- Prior knowledge of the Milgram study and being told the original results had little impact on participants’ self-predictions, suggesting that simply knowing about social pressure doesn’t help people internalize how it might affect their own behavior.
COLUMBUS, Ohio — You’re probably not as morally courageous as you think you are. Decades of psychology research have shown that most people will hurt others when an authority figure demands it. Yet a new study reveals that even when we know this disturbing fact, we still believe we’re somehow different, that we’d be the ones to walk away when others wouldn’t.
But studies like the famous 1963 Milgram obedience experiments show that’s likely wishful thinking. A staggering 65% of participants continued delivering what they believed were potentially lethal shocks to strangers, simply because an authority figure told them to.
Now, a new study published in Current Psychology reveals that even when people know about these disturbing results, they still can’t accurately predict their own behavior. The research suggests we’re remarkably bad at recognizing our own vulnerability to authority, a potential blind spot in our daily lives.
“Social pressures are way more powerful and impactful than we give them credit for,” says lead author Philip Mazzocco from The Ohio State University, in a statement. “If you fall under the sway of these pressures, you could end up engaging in behavior inconsistent with your values and morals.”
Mazzocco and his research team wanted to understand whether people actually learn from one of psychology’s most famous findings. They presented 414 adults with a detailed, first-person account of the Milgram scenario and asked them to predict how they would respond. Half the participants were even told the shocking truth about the original results beforehand.
On average, participants predicted they would quit the experiment around the 7th shock level out of 30 possible levels. They thought the “average person” would hold out until about the 12th level. But in the actual Milgram studies, most people went all the way to the end.
This held true even for people who had prior knowledge of the Milgram experiments. Psychology students across the country learn about these studies, yet this education doesn’t seem to translate into better self-awareness about their own susceptibility to authority pressure.
People’s Predictions About Obedience
Mazzocco’s team recruited participants through Amazon’s Mechanical Turk platform, paying them $1 each to complete the study. After removing participants who completed the study too quickly or didn’t follow instructions, they ended up with the sample of 414 people. The average participant was 44 years old, with 48% identifying as male and 50% as female. About 75% were white, and most had some college education. Importantly, 65% had never heard of the Milgram experiment before.
Participants read a detailed, 561-word narrative that put them in the shoes of a Milgram study participant. The scenario described every detail: being assigned the role of “teacher,” watching the “learner” get strapped into an electric chair, receiving a sample shock themselves, and then being ordered to deliver increasingly powerful shocks despite the learner’s protests.
The narrative included chilling details that mirror the original experiments: “If and when you tell the experimenter you want to stop, he says to you, ‘Please continue.’ You continue on shocking the learner until once again you say you do not want to continue. The experimenter replies, ‘The experiment requires you to continue.'”
After reading this scenario, participants were shown a photo of the actual voltage machine used in Milgram’s studies and asked to predict at which level they would refuse to continue.
The Better-Than-Average Effect Clouds Our Judgment
The study revealed a psychological phenomenon that explains why we’re so bad at predicting our own behavior. Researchers call it the “better-than-average effect,” our tendency to view ourselves more favorably than others. Participants consistently predicted they would be more likely to disobey than the average person, even when confronted with hard evidence about how people actually behave.
The researchers found that most adults either can’t or won’t apply social psychology research, especially the Milgram obedience results, to how they think they would behave themselves.
The researchers also discovered that telling people about the original 65% obedience rate had a limited impact on their predictions. While it did make participants predict slightly more obedience, both for themselves and others, the effect was much stronger for their predictions about other people than for themselves.
This pattern points to a broader problem in how we process information about human behavior. People tend to underestimate the power of situations and overestimate the importance of individual character. We think we’re different, that our personal values and moral compass would guide us to do the right thing, even when presented with clear evidence that situational pressures often override individual differences.
Women predicted they would be less obedient than men, and older participants predicted less obedience for both themselves and others. People who scored higher on personality traits like conscientiousness and openness also predicted they would be more likely to disobey.
Previous research has shown that in actual obedience experiments, conscientious people are more likely to obey authority figures, not less. This suggests that people who think of themselves as rule-followers imagine they would follow the rule “do no harm,” when in reality they might follow the rule “obey the experimenter.”
The researchers note that it’s uncertain whether students or the general public are truly absorbing this research, especially when it comes to understanding human behavior and reflecting on their own sense of self.
Consider situations where someone in authority asks you to do something that makes you uncomfortable: workplace harassment, military settings, corporate misconduct, or even everyday scenarios. If we can’t accurately predict our own behavior in these situations, we’re less likely to prepare for them or recognize when we’re being influenced.
Decades after Stanley Milgram shocked the world with his findings about obedience to authority, we still haven’t truly internalized the lesson. We continue to believe we’re somehow different, somehow better, somehow more moral than the average person who would blindly follow orders.
Paper Summary
Methodology
Researchers recruited 456 participants through Amazon’s Mechanical Turk platform and had them read a detailed, first-person narrative of the Milgram obedience experiment. After excluding participants who completed the study too quickly, took too long, didn’t answer key questions, or requested removal during debriefing, 414 participants remained. Participants were randomly assigned to two groups: half were told that 65% of people in the original study showed complete obedience, while the other half received no information about the results. All participants then predicted at what shock level (1-30) they would quit the study, and made a separate prediction for the “average person.” The study also measured participants’ personality traits, demographics, and prior knowledge of the Milgram experiments.
Results
Participants dramatically underestimated obedience levels, predicting they would quit around shock level 7 while predicting others would quit around level 12. This pattern held true regardless of whether participants knew about the original Milgram results or had prior knowledge of the experiments. The study confirmed a “better-than-average effect”—people consistently predicted they would be more likely to disobey than others. Women and older participants predicted less obedience for both themselves and others, while people higher in conscientiousness and openness predicted they would be more likely to disobey.
Limitations
The study used imagined scenarios rather than real obedience situations, which may not capture the actual psychological pressures present in live experiments. The sample was recruited online and was predominantly white and well-educated, potentially limiting generalizability. Some of the significant findings had small effect sizes, and the researchers noted that future studies should use preregistration and power analyses for more robust methodology.
Funding and Disclosures
This research was supported by a small grant from the Ohio State Mansfield Student Research fund. The authors reported no conflicts of interest, financial or otherwise. The study was reviewed and approved by the Ohio State University Office of Responsible Research Practices and adhered to ethical standards of both the university and the American Psychological Association.
Publication Information
The study “Milgram shock-study imaginal replication: how far do you think you would go?” by Philip J. Mazzocco and colleagues was published in Current Psychology in 2025. The paper was accepted on May 13, 2025, and published online on May 29, 2025.
