Study guides › Social psychology
Milgram’s Obedience Study, Explained for the Exam
The classic study you will be asked about again and again. Learn what Milgram did, the numbers examiners want, and exactly how to turn it into top-band answers.
The 30-second version
In 1961, Stanley Milgram tested whether ordinary people would obey an authority figure who told them to hurt someone. 40 men were told to give a “learner” electric shocks for wrong answers, rising in small steps from 15 to 450 volts. The shocks were fake, but the participants didn’t know that. 26 out of 40 (65%) went all the way to 450 volts. Every single one reached 300 volts.
Milgram in 90 secondsHand-drawn video coming soon
Why this study is worth your time
Milgram (1963) is the classic study for obedience in Edexcel Social Psychology, on both the IAL (WPS01) and A Level (9PS0) courses. You can be asked to describe it, evaluate it, compare it with other studies or use it in an essay. Learn it properly once and it pays you back in almost every Social question.
Examiners reward two things:
- AO1 – knowledge. Precise detail. Numbers, not vibes: “65% went to 450V” beats “most people obeyed”.
- AO3 – evaluation. A judgement backed by evidence and explained to the end. “It was unethical” on its own scores almost nothing.
Everything below is built around those two.
Why did Milgram do it? (Aim)
After the Second World War, people were asking a disturbing question about the Holocaust: did the people who carried it out do so because they were unusually cruel, or because they were following orders? Milgram wanted to find out how far ordinary people would go.
Aim: to investigate how far ordinary people would obey an authority figure who told them to harm another person.
Exam tip: if you’re asked for the aim, write it as one clear sentence. There are no marks for the history lesson.
What did he do? (Procedure)
- Recruitment. 40 men aged 20–50 from the New Haven area answered a newspaper advert for a “memory study” at Yale University. They were paid $4.50 just for turning up.
- A rigged draw. Each participant met “Mr Wallace”, a friendly 47-year-old man who was really an actor (a confederate). The draw was fixed so the real participant was always the teacher and Mr Wallace was always the learner.
- The set-up. The learner was strapped into a chair in the next room with an electrode on his wrist. The teacher was given a real 45-volt sample shock, so the machine felt real.
- The task. The teacher tested the learner on word pairs. Every wrong answer meant a shock, and each shock was 15 volts higher than the last.
- The pressure. An experimenter in a grey lab coat stayed in the room. If the teacher hesitated, he used four standard prods, always in the same order.
The shock generator: 30 switches in 15-volt steps, labelled from “Slight Shock” up to “Danger: Severe Shock”, with the last two simply marked “XXX”.
At 300 volts, the learner pounded on the wall. At 315 volts he pounded again – and then went completely silent. Silence counted as a wrong answer, so the teacher was told to carry on.
The four prods
Prod 1“Please continue” or “Please go on.”
Prod 2“The experiment requires that you continue.”
Prod 3“It is absolutely essential that you continue.”
Prod 4“You have no other choice, you must go on.”
If the teacher still refused after prod 4, the experiment stopped.
Remember it like this
The staircase you can’t find the top of. Each shock is only one small step higher than the last, so there’s never an obvious moment to say “this is where I stop”.
What did he find? (Results)
65%went all the way to 450V (26 of 40)
100%reached at least 300V
12.5%stopped at exactly 300V (5 of 40)
0–3%what Yale psychology students predicted would reach 450V
Participants didn’t obey happily. Many showed extreme stress: sweating, trembling, stuttering, biting their lips and digging their nails into their hands. 14 showed nervous laughter, and three had full-blown, uncontrollable seizures.
Afterwards, every participant was debriefed: they met Mr Wallace again, unharmed, and were reassured. In a follow-up questionnaire, 84% said they were glad to have taken part.
Exam tip: learn three numbers – 40 participants, 65% to 450V, 100% to 300V. They are the fastest AO1 marks in the whole topic.
What does it mean? (Conclusions)
Ordinary people will obey orders to harm another person, even against their conscience, when the authority seems legitimate. Milgram concluded the cause was the situation, not the person: a famous university, a calm experimenter in a lab coat and tiny 15-volt steps all made it hard to say no.
Link it to agency theory
Milgram later explained this with agency theory. Normally we are in an autonomous state and feel responsible for what we do. Under an authority figure we can shift into an agentic state: we see ourselves as carrying out someone else’s wishes, so the responsibility feels like theirs. The stress participants showed is called moral strain – knowing something is wrong but feeling unable to stop.
Common trap
Milgram did not prove people are cruel or evil. He didn’t test personality at all. He tested a situation and showed how powerful it can be.
How to evaluate it (AO3)
Use Point – Prove – Consequence for every evaluation: make the point, prove it with evidence, then explain what it means for Milgram’s conclusions. Here are five you can use.
Strength
It can be replicated
PointMilgram used a highly standardised procedure.
ProveEvery participant had the same learner responses, the same four prods in the same order and the same 15-volt steps.
ConsequenceOther researchers can repeat it to check reliability. Burger (2009) ran a partial replication, stopping at 150V for ethical reasons, and found similar levels of obedience, which suggests the findings are reliable.
Strength
It matters in the real world
PointThe findings help explain real obedience to harmful orders.
ProveHofling et al. (1966) found that 21 out of 22 nurses were prepared to give a dangerous dose of a drug when a “doctor” told them to over the phone.
ConsequenceObedience to authority is not just a lab effect, which supports Milgram’s conclusion and makes the study useful for understanding places like hospitals and the military.
Weakness
Did they believe it?
PointParticipants may not have believed the shocks were real.
ProveOrne and Holland (1968) argued that participants trust experimenters not to let anyone get hurt. Perry (2013) listened to Milgram’s original recordings and found that many participants voiced doubts.
ConsequenceIf participants were playing along, the study may measure trust in the experiment rather than obedience, which lowers its internal validity.
But: the severe stress many participants showed suggests a lot of them did believe it.
Weakness
A narrow sample
PointAll 40 participants were American men.
ProveThey were volunteers from one area answering an advert, and no women were tested in the baseline study.
ConsequenceThe results may not generalise to women or to other cultures.
But: Burger (2009) found no significant difference in obedience between men and women.
Weakness
Serious ethical problems
PointThe study broke key ethical guidelines.
ProveParticipants were deceived (a “memory study”, a rigged draw, fake shocks), the prods made it hard to use the right to withdraw, and three had seizures, so they were not protected from harm.
ConsequenceThe study could not be run today in its original form.
But: Milgram didn’t expect such high obedience, everyone was debriefed, and 84% said they were glad to have taken part.
Exam tip: for full marks on an “evaluate” question, finish with a judgement: which point matters most, and why?
Mistakes that cost students marks
“The learner screamed in pain.”
In the 1963 baseline, the learner only pounded on the wall at 300V and 315V, then went silent. The screams belong to Milgram’s later variations.
“Participants were given electric shocks.”
The learner supposedly got the shocks, and they were fake. The only real shock was the teacher’s 45V sample.
“Most people obeyed.”
“65% (26 of 40) went to the maximum 450V.”
“It was unethical.”
Name the guideline (deception, right to withdraw, protection from harm), give the evidence, then weigh it up.
Model answer: 8 marks
Practice questionEvaluate Milgram’s (1963) study of obedience. (8 marks)
Edexcel 8-mark questions are marked in levels. They reward accurate knowledge (AO1) and balanced evaluation that reaches a judgement (AO3). Check the exact split in your specification’s sample mark scheme.
AO1 knowledge AO3 evaluation
Milgram (1963) studied 40 American men aged 20–50, who were told to give increasing electric shocks from 15V to 450V to a learner whenever he made a mistake. Although the shocks were fake, 65% of participants went all the way to 450V. A strength is that the procedure was highly standardised, as every participant received the same four verbal prods and the same 15-volt increases. This means the study can be replicated to check reliability, and Burger (2009) found similar obedience in a partial replication to 150V, suggesting the findings are reliable.
However, the sample was only American men, so the findings may not apply to women or to other cultures, although Burger’s finding of no significant gender difference weakens this criticism.
A further weakness is that participants may not have believed the shocks were real. Perry (2013) found that many participants expressed doubts in Milgram’s original recordings, which suggests the study may have measured trust in the experimenter rather than genuine obedience. On the other hand, many participants sweated and trembled and three had seizures, which suggests they did believe they were causing harm.
Finally, the study raises serious ethical issues. Participants were deceived and the prods made it difficult to withdraw. Milgram did debrief everyone and 84% said they were glad to have taken part, but the distress caused means the study could not be repeated today.
Overall, while ethical concerns and doubts about believability limit the study, its standardised procedure and support from later replications mean it still provides strong evidence that ordinary people will obey authority figures, even when it means harming others.
Why this scores well
- It opens with precise AO1 in two sentences, then moves on.
- Every evaluation point is explained through to its consequence.
- It uses other studies (Burger, Perry) as evidence.
- It ends with a judgement, not a summary.
Test yourself
Answer each one out loud before you click. If you can do all six without looking, you know this study.
1. How many participants were there, and who were they?
40 American men aged 20–50 from the New Haven area, recruited by a newspaper advert.
2. What percentage went to 450V?
65% (26 out of 40).
3. What happened at 300V?
The learner pounded on the wall. Every participant reached this point, and 5 (12.5%) stopped there.
4. What was the second prod?
“The experiment requires that you continue.”
5. Name three ethical issues with the study.
Deception, the right to withdraw and protection from harm.
6. What is the agentic state?
When a person sees themselves as acting for an authority figure, so they feel less responsible for their own actions.
For teachers
Teaching this next week?
The PsychHack IAL Social Psychology pack turns this study into a ready-to-teach lesson with slides, a student workbook and exam practice. It launches soon – join the newsletter above to hear first.