Milgram Experiment (Video)

0:02 Introduction

Today we'll learn about the Milgram experiment. The Milgram experiment was a set of studies at Yale University in the early 1960s. An experimenter told participants to give what they believed were increasing electric shocks to another person, and in the first study 65 percent of participants went to the maximum. The video has nine parts.

0:27 The question Milgram asked

First, we will look at the question Milgram asked. Stanley Milgram was a psychologist at Yale University. He began his experiments in the summer of 1961, a few months after the trial of Adolf Eichmann began in Jerusalem. His question concerned obedience, the following of instructions from a person seen as an authority. He asked whether an authority could direct ordinary people to take part in actions that harm another person. Milgram had in mind the Holocaust, in which many ordinary people took part. He reported the first results in 1963, in the Journal of Abnormal and Social Psychology. In the studies, a participant from the New Haven area sat at a shock generator in a Yale laboratory and followed the instructions of an experimenter in a laboratory coat.

1:18 A common misreading

Second, we will correct a common misreading of the studies. Three claims are common: that people in general will obey an authority to the point of harm, that participants felt nothing, and that the studies produced one fixed percentage. In the first study, 65 percent of participants went to the end, and 35 percent stopped earlier. Participants showed visible strain, so the claim that they felt nothing is false. Across the variations, the share of participants who went to the end ran from 0 to 92.5 percent, so the features of the situation changed the rate.

1:58 The procedure

Third, we will go through the procedure, so that you can see where these figures come from. Milgram recruited 40 men aged about 20 to 50 through newspaper advertisements, and none of them studied at Yale. Two men arrived for each session, the participant and a confederate, an actor who plays a role in a study without the participant knowing. The two men drew slips of paper to decide who would be the teacher and who would be the learner. Milgram rigged the draw, so the participant was always the teacher. The experimenter strapped the learner into a chair in another room. The teacher read word pairs aloud and gave a shock for each wrong answer, raising the level by 15 volts each time. The generator had 30 switches, labeled from 15 volts to 450 volts. The learner received no real shocks, and Milgram had planned the learner's protests in advance. At 300 volts the learner pounded on the wall and stopped answering. He pounded again at 315 volts, and after that he made no sound. When the teacher hesitated, the experimenter used four standard prods in order. The prods were please continue, the experiment requires that you continue, it is absolutely essential that you continue, and you have no other choice, you must go on. After the session, the participant met the learner and was told that the shocks had caused no harm. Milgram mailed a written report on the purpose of the studies after the series ended.

3:33 The results

Fourth, we will go through the results. The best-known figures come from the first published study of 40 men, in which the learner sat in another room. All 40 participants went to 300 volts, the point at which the learner pounded on the wall, and five of them stopped there. Twenty-six of the 40, or 65 percent, went to the maximum of 450 volts. The other 14 stopped at levels between 300 and 375 volts. Before the study, Milgram asked people to predict the outcome, and they expected that only a tiny minority would go to the end. Many participants sweated, trembled, stuttered and bit their lips, and 14 of the 40 laughed nervously, in three cases in uncontrollable fits. The strain appeared among the participants who continued to the end as well as among those who stopped.

4:30 The variations

Fifth, we will compare the variations. Milgram then changed one feature of the situation at a time. When the learner sat in another room, 65 percent of participants went to the end. With the learner in the same room the figure was 40 percent, and when the teacher had to press the learner's hand onto a plate the figure was 30 82 00:04:52,909 --> 00:04:52,859 percent. The closer the learner sat to the teacher, the lower the rate of full obedience. When the experimenter gave instructions by telephone, full obedience fell to 22.5 percent. When Milgram moved the study from Yale to a run-down office building in Bridgeport, full obedience was 47.5 percent. When two other teachers refused to continue, 4 of 40 participants went to the end. When a peer delivered the shocks and the participant did a subsidiary task, 37 of 40 stayed and took part until the end. In several conditions, including one in which two experimenters gave conflicting orders, no participant went on to 450 volts. These differences point to features of the situation more than to a fixed trait of the participants.

5:45 Ethical criticism and new rules

Sixth, we will examine the ethical criticism and how the rules for research changed. In 1964 the psychologist Diana Baumrind published a critique in American Psychologist. She objected to deception, giving participants false information about the purpose or the procedure of a study. She also pointed to the distress that participants experienced. Baumrind questioned whether a debriefing could repair the loss of dignity, self-esteem and trust in rational authority that the procedure risked. She also criticized Milgram's apparent indifference to the distress he observed. Milgram replied that he had debriefed the participants. In his follow-up survey, which had a response rate of 92 percent, 84 percent said they were glad or very glad to have taken part. Critics answer that these reports do not settle the question of harm, and Gina Perry has argued that not all participants received a proper debriefing. The present US rules for research with people date from the 1970s, and Milgram's work was one case among several that shaped those rules. In 1974 the US National Research Act required institutions that receive federal funds to set up institutional review boards, committees that review research with people before the research starts. The Act followed several cases of research abuse, among them the Tuskegee syphilis study. The Belmont Report, published in 1979, set out three principles, respect for persons, beneficence and justice. Respect for persons requires voluntary and informed participation, and beneficence requires the researcher to minimize harm and maximize benefit. Justice requires a fair selection of participants. Informed consent is a participant's agreement to take part after the researcher explains what the study involves, and a participant may withdraw at any point without penalty. A review board asks whether the researcher has minimized the risks and whether the benefits outweigh the risks. Where a study needs deception, the researcher has to justify the deception, keep the deception to a minimum and debrief participants fully at the end. Review boards today are unlikely to approve the original Milgram design.

8:12 Replications and later critiques

Seventh, we will study the replications and the later critiques of how the results were interpreted. Because review boards are unlikely to approve a full replication, researchers built a safer version of the procedure to test whether the result still holds. In 2006 Jerry Burger ran a partial replication with 70 adults, and he published the study in 2009. Burger used Milgram's later version, in which the teacher could hear the learner through the wall, and he stopped each session at 150 volts, the point at which the learner first demanded to be let out. He chose 150 volts because 79 percent of Milgram's participants who went past that point continued to the end of the range. A participant's decision at 150 volts therefore predicts the later decision. Burger screened participants for psychological problems, gave a short debriefing, and told participants that they could withdraw and keep their payment. Obedience was only slightly lower than in Milgram's studies 45 years earlier, and women and men did not differ. In 2015 Dolinski and colleagues ran a similar procedure, with levels up to 150 volts, with 80 participants in Poland, and they published the study in 2017. They reported high obedience and found no clear differences by sex. A stop at 150 volts estimates the later behavior and does not measure that behavior directly. The replications support the finding, and the argument among researchers now concerns why people continued. Milgram's own account was that participants entered an agentic state, in which they saw themselves as carrying out another person's wishes. Haslam, Reicher and colleagues examined the prods. In Milgram's sessions, the prod that read most like an order was the least successful, and prods that appealed to scientific goals were more successful. The same group later tested the prods in an online analog of the procedure. Continuation rose as a prod appealed more to scientific goals, and fell as the prod became more like an order. They propose engaged followership, in which people continue because they identify with the goals of the leader, here the scientific project. A second critique concerns belief in the cover story. Perry and colleagues reanalyzed an unpublished study by Milgram's assistant on whether participants believed the shocks were real. In 18 of 23 variations, participants who fully believed that the learner was in pain gave lower shocks on average than participants who doubted the cover story. Perry's reading of Milgram's archive also suggests that the sessions varied more than the published account states, which limits comparisons between the variations. These critiques mainly concern why participants continued, and the differences between the variations remain in the published data, within the limits that Perry describes.

11:27 What the studies can and cannot show

Eighth, we will separate what the studies can show from what they cannot show. The studies show that many ordinary participants continued under pressure from an authority, and that the rate depended on the situation. They do not give a single percentage for people in general, because the participants were not a representative sample. Each condition had about 40 participants from one part of Connecticut, and the figure changed widely between conditions. The studies also do not support direct claims about real atrocities, which unfold over months or years and involve motives that a one-hour laboratory session cannot reproduce. Deception limits the result as well. Some participants doubted the cover story, so their behavior was not a pure test of obedience.

12:18 What researchers do today

In the final section, we will learn what researchers do today. A researcher who studies behavior under pressure follows a fixed set of steps before and during the study. You submit the protocol to an ethics board before you recruit participants, and the board decides whether the study may begin. You obtain informed consent from each participant, explain the right to withdraw and give your contact details. The board also checks that you select participants fairly and that you have reduced the risks as far as the design allows. If your study needs deception, you justify the deception in the protocol and plan a full debriefing. You decide in advance how to respond to distress, and you stop a session if a participant's distress rises. The board weighs the risk to participants against the knowledge the study is expected to produce, and you keep a record of the board's approval.

13:15 Summary and question

Here is a short review of the main points. In Milgram's first study, 65 percent of participants went to the maximum of 450 volts. The variations showed that features of the situation moved the rate from 0 to 92.5 percent. Criticism of the deception and distress, together with cases of abuse such as the Tuskegee study, shaped the review boards, consent rules and debriefing that research uses today. Partial replications that stop at 150 volts still find high obedience, and later critiques dispute the explanation of why participants continued. Test what you learned with this question. Why did Jerry Burger stop his replication at 150 volts, and what does his approach assume? Pause the video and work it out. The answer follows. Burger stopped at 150 volts because the learner first demands to be let out at that point, and 79 percent of Milgram's participants who went past 150 volts continued to the end of the range. The approach assumes that today's participants who continue past 150 volts would behave as Milgram's participants did, so the result estimates the later behavior and does not measure that behavior directly.