The Science of Everything Podcast - Episode 33: Disturbing Social Psychology Experiments

Episode Date: June 22, 2012

A discussion of three of the most chilling experiments in the field of social psychology: the Ash Conformity Experiment, the Stanford Prison Experiment, and the Milgram Obedience Experiment. In each case I discuss the motivation and setup of the experiment, outline the results, discuss replications and variations of the original experiment, and end with a look at the implications of the experiment for understanding the darker side of human nature.

You're listening to The Science of Everything podcast, episode 33. Disturbing Social Psychology Experiments. And I'm your host, James Fodor. In this episode, we're going to look at three very famous classic experiments in social psychology, the Ash Conformity Experiment, the Stanford Prison Experiment, and the Milgram Obedience Experiment. I'm going to explain the experiments, explain what they're about and how they were conducted, take a look at the results and see perhaps the surprising nature of said results. Then I'll discuss some replications.

and extensions of these original studies that were done later on to expound and test the results. And I'll take a look at an analysis of the experiments to discuss their validity and have a look at some real-world applications of what we can learn from these experiments. I've grouped all these three together because although they study somewhat different things, they're all, I think, quite disturbing in their implications they have about perhaps human nature or human interactions. and very classic, as I said, experiments to study in social psychology courses and the like. So now that I've wet your appetite, let's start with the Ash Conformity Experiment. So this experiment was conducted in 1951 by Solomon Ash.

The basic idea of the experiment was that he took a bunch of college students, seven or nine. Each of the groups of students were shown a card with four lines on it. One was a reference line, and then there were three other lines sort of next to it. and basically all you had to do was pick which of the three lines was the same length as the reference line. Now, I mean, this could be difficult if the lines were pretty close, but the experiment was deliberately done so that the lines were, it was very obvious which answer was correct.

I mean, one was like five centimetres longer, one five centimeters longer, one five centimeters shorter. You can look this up if you want to check the actual originals for yourselves, but it was very obvious. In controlled versions where they just had one subject doing this by themselves, they got it right like 99% of the time, so it was really easy. So, as I said, they did this, they picked the right line, or what they thought was the right line, in groups of around eight or so.

However, what they didn't know was that in groups of nine, eight of the subjects were actually confederates of the experimental, you know, the group size varied a bit, but every subject in the group except for one was actually a confederate of the experimenter, which means that they were in on it. And in some of the trials, the Confederates had been instructed to deliberately make ridiculous judgments or incorrect judgments, or incorrect judgments about which line was correct. But to unanimously agree with each other, so in some of the trials they were told you picked number two, even though that's clearly not the right answer,

or pick number three or whatever. The point of the study was to see whether the one other subject, who wasn't a Confederate, would go along with the group or not? Would they go along with whatever else was saying, even though it was clearly wrong? Would they conform, in other words? Or would they go with what they thought was correct? This is why this is called the conformity study,

because it was a study of whether people would conform to a group norm or not. So as I said, 99% of the time, when people were just doing this by themselves, they made the correct judgment. So how often, or how much do you think people would conform with the wrong answer when eight other people were saying that it was true, even though you clearly saw that it wasn't? Well, the answer was, on average, about one-third of the time. The average conformity rate across all the trials was 33%. A total of 75% of the subject, so this is the people who aren't in on the experiment, 75% of them conformed to the wrong answer at least. least once, and 50% of them conformed on more than half of the trials that they were involved in.

Interesting, only 5% always conformed on every single trial. So most people conformed at least conformed some of the time, but not all of the time, which is somewhat interesting in and of itself. But the overall rate of conformity was 33%. So Ash fiddled around with this a bit to see how much of an effect he could get with different variations in the length of the lines. he found basically the same effects of conformity when he got the subjects to declare that a basically around 30 centimetre line was equivalent to a 10 centimeter line so basically he got people conforming 33% of time to say that a 10 centimeter line was the same length as a line three times as long which is obviously you know not true so people are a substantial portion of the

time willing to say things that are flatly contradictory to basic sensory impressions. Now, an interesting part of this experiment was Ash asked people why they made such obviously incorrect judgments after the experiment. So some of the reasons that they gave were that, well, they thought they must have been looking at the line widths or something like that. They thought they must have been doing something wrong. Or I just assumed it, or they said, I just assumed it was an optical illusion. Or comments like, if eight out of nine people made the same choice, I must have missed something

in the instructions or it must have done something wrong. Basically, that's what most people thought. They thought that if they were getting it wrong on such an obvious, easy task, they must have missed something. And they were essentially using the information provided by the selections of the other people to infer something about the situation that they had not otherwise picked up or that, you know, they thought they hadn't picked up. And pretty much everyone in this experiment felt the reported feeling anxious

and feared disapproval from others, and was very self-conscious of the situation, even if they didn't conform. And they're often, when they didn't conform, when they said they disagreed, like they were often very highly apologetic about disagreeing and clearly embarrassed by it. So you can see photos.

I've seen photos of this experiment. It's immediately easy, at a glance, to see which of the people in the trial is the actual subject and which are the confederates, because the subject looks just very discompeted and anxious and concerned about what's going, on, whereas everyone else is just perfectly relaxed. So you can see that just in this basic experiment,

people get very anxious and concerned when they are asked to potentially, you know, disagree on a seemingly simple issue with a bunch of their peers. So this experiment has been replicated in a bunch of different ways, some by Ash and some by other experimenters. Here are some of the variations that have been done. One important alteration was the presence of a dissenter in the group, that is one other person who was picking the right choice rather than deliberately picking the wrong one like previously all of the Confederates had done. He found that the presence of only a single dissenter

among the group of Confederates was enough to pretty much completely get rid of the conformity effect. So if there were nine people and eight of them all agreed with themselves, then the conformity effect was high. But if only one of them disagreed, then pretty much the conformity effect completely went away. So the presence of dissent seems to be very powerful in social situations like this. Another rather interesting effect is the attractiveness of other members in the group.

The conformity effect was increased when the attractiveness of the other members of the group was higher. Another factor that was varied was the complexity or the difficulty of the task that was being done. Now, in the original case, it was just judging the length of lines, which was pretty easy task. The more difficult you make the task, the higher the level of conformity is, which sort of makes sense, because you expect people to be more likely to go with the crowd if they're less sure of themselves. Another factor was the cohesiveness of the group, so people tend to conform more if friendships or mutual dependencies within the group were set up beforehand. So if there were people that they knew, they were more likely to conform than if they were complete strangers, or people

that they felt some sort of connection to. Another important variation in the experiment was that in the original experiment, people had to publicly announce their answer. I should have said this before, actually. What they did was they went around the circle and the actual subject, the only non-Confederate, was one of the last to announce their answer. So they got to hear that everyone else was agreeing with the wrong answer. And that's what put a lot of the pressure on for them to conform. When you altered the experiment so that the subject just had to write down their answer instead of calling it out, so everyone else called theirs out, but the subject was allowed to write it down because obviously you had to do that otherwise, the subject wouldn't know that everyone else was picking a different answer.

conformity was reduced to an average of about 12.5%, which is substantially less than the 33% in the original experiment. Although it's still surprisingly high, when you're given that the issue of looking stupid is out of the question because they're anonymously writing down their answer, still relatively high rates of conformity. And a later study by Deutsche and Gerard found average rates of conformity of 23%, even with high anonymity in the conditions of responding,

and high certainty about the answer. So that would be very obvious difference in line lengths and, high anonymity in terms of who is who knows what your answer is given. So it seems there are multiple effects going on here. One is that people are feeling anxious and disconcerted at disagreeing with the rest of the group and so they're going along with the crowd even though they think

that the rest of the group is wrong. That's part of the effect because when you increase anonymity the rates of conformity fall. But they also seems to be part of an effect which is informational which is that people are using the fact that other people disagreeing with them in an apparently simple task to infer that they must have done something wrong or misunderstood something and so that they actually think that they are wrong, which is, in some sense, rational actually. Those who tend to conform typically are people who have high

anxiety and low social status and need for approval from others. One very interesting replication of this, well, that's not exactly replication, but it's another experiment which was conducted by Milgram, the same guy from the Milgram Obedience Experiment, which we'll talk about later on in this episode, found that if one individual stops and stares at the sky on a busy street, so just start looking up and 4% of people would stop as well, and 40% of people would look. So this is just random passes by. 40% would look up, but only 4% would stop. However, if 15 people stop and look up to the sky,

the numbers increase to 40% of passes by stopping, as opposed to 4% before, and 90% of people looking up at the sky, as opposed to 40% before. So once again, that's kind of not surprising. I've actually seen data where they have groups of all different sizes looking up at the sky and you see the rates of passes by stopping, looking up, dramatically increase as you get more than like two or three. It doesn't take very many. But the key thing seems to be like more than one or more than two people doing something and other people immediately start conforming. There's a lot of evolutionary psychological basis for this sort of thing. Homersapians being a

social species and there being elements of almost hardwired in a sense desire to or tendency for humans to conform to the behavior of those around them. You can see this in, for example, socialization of children, picking up of accents, tendency for customs or social norms to become strongly ingrained in a society or in a culture. These are things we'll look at in future episodes, by the way. Okay, so that's the Ash Conformity Experiment. You might not have found that terribly disturbing, but I deliberately put it first because it's probably the least disturbing of the three, although I found it somewhat disturbing, that people will essentially say things that apparently patently wrong, just because other people are. Now we'll move on to the Stanford Prison

Experiment, which is quite famous, actually. You may have heard of it, but there's a little bit of misinformation out there about what exactly happened, so I'll try and break through some of that. So the Stanford Prison Experiment was held in 1971 at Stanford University, as the name would indicate, by a team of researchers led by the psychology professor Philip Zimbardo. So this Zimbardo was rather an interesting character. What they did was they got 24 males students. So once again, this experiment was on the students. These 24 male students were judged to be psychologically and physically healthy and stable, so they deliberately tried to pick, you know, fairly normal and stable people. And so they get these 24 stable male students and they were,

and they were randomly assigned to roles of either prisoners or guards in a mock prison situation, which was conducted in the basement of the psychology building at the university. So it was basically, it was called a prison experiment because it was a mock prison environment. The participants in the experiment were mostly white middle class, and as I said, there were male college students. And the group was deliberately selected to avoid those, to exclude people with criminal background or psychological impairments or anything else like that. So they were deliberately picked to try and be stable individuals and ordinary individuals and then randomly assigned to either prisoners or guards. Okay, so the instructions, one of the controversies in this experiment is the degree to which people were play acting deliberately, and the degree to which they had truly internalized their roles.

And you'll see what I mean by that in a minute. But the people who were randomly assigned to be guards were given wooden batons and, I believe, types of some type of uniform, to establish their status. Yeah, clothing similar to the actual prison guards. They were given uniforms and also mirrored sunglasses to prevent eye contact with the prisoners. So they were essentially given clothing and props to represent their status. Prisoners were likewise fitted with uncomfortable smocks and stocking caps, a chain around one ankle, and they were also assigned numbers that were sewn onto their uniforms.

And guards were instructed to call prisoners by their numbers. rather than their names, so as to essentially dehumanize them and treat them more like prisoners and so on. The prisoners were actually full-on arrested in their homes and taken to the university in a police car. They were charged, well, pretend to charge with armed robbery. Their fingers were taken, their mugshots were taken, and they were transported from the police station where they were actually taken to the Mock prison in the university and given their numbers and smocks and everything else. So in the Mock University there was a number of cells, a prison yard,

a smaller cell for solitary confinement and a few other rooms, like for eating and someone like that. But, yeah, it was a mock prison setup. The prisoners were supposed to stay in their cells all day and night until the end of the study, which I think were supposed to last for like a week or two weeks or something. The guards worked in teams of three for eight-hour shifts. So the guards didn't actually have to stay on during the night or after their shift during the experiment. So only some of the guards were there at any one time, and the rest of them were off doing whatever they were doing.

whereas the prisoners had to stay there all the time. So that is an important difference. A number of the rules that were given was that the prisoners were supposed to follow the orders of the guards, and they were supposed to ask permission to do anything, even going to the toilet or eating or anything like that. So there was a definite effort to try and make it feel like a prison environment and to make the prisoners feel like they were subjected to the authority of the guards.

However, the guards were instructed not to physically harm the prisoners, but they were told to keep order and enforce the rules. and they were allowed to make life uncomfortable for the prisoners if this was helpful to enforcing the rules. So they weren't allowed to physically harm the prisoners, but they were allowed to make life uncomfortable for them. That was the instructions they were given. Okay, so that's the setup. We got a mock prison environment, Stanford University, randomly selected bunch of apparently normal and stable male students who are being put in these roles of prisoners or guards.

Okay, so before too long the prison guards, by all accounts, started acting rather mean towards the prisoners. So some examples of things that they asked them to do. They decide to call the role during the middle of the night, make the prisoners do push-ups, with the guards putting their feet on the middle of the prisoners back during push-ups, which is obviously not going to make the job any easier. On the second day of the experiment, the guards crushed a rebellion of the prisoners

and became more verbally abusive towards them. After 36 hours, one prisoner apparently began to act crazy, as Zimbardo described it. He began to scream and shout and curse and seem out of control. Eventually, he had to be released from the experiment, sanitary conditions apparently declined quite quickly as the guards refused to allow some of the prisoners to go to the toilet anywhere but a bucket which was in their cell so that's kind of gross

mattresses were given to the prisoners but the guards punished their prisoners by confiscating these and leading them to sleep on the concrete floor several of the guards became more cruel verbally abusing the prisoners as I said exhibiting rather sadistic tendencies and harsh punishments like as I said taking their um taking their mattresses or or their food, or putting them in solitary confinement, things like that. Apparently on the fifth day of the experiment, all of the volunteer prisoners, well, I mean all of them were volunteers, but all of the prisoners asked to be released from the experiment, even though that would mean forfeiting their pay.

So at this point, I think it was like five or six days, yeah, six days the experiment had gone on. It was sort of ambiguous as to whether the prisoners were allowed to leave or not, because they were there voluntarily. I mean, they weren't really arrested. As far as I know, when they asked to leave, I don't exactly know what happened. it's hard to get detailed reports of this sort of thing, but they, well, they didn't leave anyway. So it was a bit ambiguous, so was this real or was this a play act? Were they really prisoners? Could they really leave if they wanted to?

This was one of the issues of the experiment, which we'll come back to later on. Eventually, Zimbardo aborted the experiment after six days of the plan two weeks, when basically a graduate student sort of saw the conditions of the experiment and objected and said, oh, basically, this is getting out of hand here. to, we need to end this now. And Zimbardo said that of more than 50 people who had observed the experiment, I don't exactly know who these people were, but presumably faculty members, maybe some of the police and other individuals, this graduate student was the only one who questioned the morality or the sense of this experiment. And so, as a result of this concern

being raised, the experiment was discontinued after only six of the planned 14 days. What's the real point of this experiment? Why is it such an issue? Well, the main issue is because it seems to indicate that people will do bad things to other people if you put them in a situation of power over them. This is the basic idea that power corrupts in absolute power corrupts absolutely. It seems that this experiment sort of confirms that sort of finding that fairly ordinary people and stable individuals could become quite mean and sadistic given the right environment. I haven't exactly done justice to the tale properly in this episode. It might be worthwhile to go and read accounts of this to get a bit. bit more of a feel as to the sort of the details of the situation and how the guards treated the

prisons and so on. No, as I could, nothing along the lines of physical abuse that I could tell, but a lot of psychological abuse and sadistic, nasty treatment and verbal abuse. Even though they weren't instructed explicitly to do any of those things, they were just instructed to keep order. So, you might wonder how representative these findings are of, you know, real-world situations or how valid are they outside of the confines of this original experiment? And as I've said, there were a number of criticisms of this experiment.

One is that Zimbardo himself took on the role of the overseer or essentially the boss of the prison. So he himself was involved in his own experiment, which is methodologically problematic, so he potentially could have influenced the results in some ways, subtle or not so subtle ways. Another criticism is that, as I said, people were just play-acting. They knew it wasn't real, or it. some level knew it wasn't real, so they just played the roles that they thought were expected of them. Zimbardo countered this by saying, well, even if there was some degree of role playing initially, quite quickly, the prisoners and guards internalized their roles. That is, sort of

really came to identify with the positions that they were occupying and the situation that they were in, and so it quickly evolved into more than just role playing. Another concern was that there may be in selection bias in terms of the people involved in the experiment. So, because of this concern, some researchers at Kentucky University, sorry, Western Kentucky University recruited students for a study using an advertisement similar to one to the one used in the original experiment, both with and without the words prison life appearing in the advertisement. So this is the key difference, whether the words prison life were or not included in the advertisement. And they found that students volunteering for a study with

prison life in the title possessed greater dispositions towards abusive behavior than those who volunteered for the one without those words in the study. I don't know exactly how they measured that, I presume, from some sort of personality test or something. I'd have to check those details. But basically, there does seem to be some evidence that there may have been a selection effect in terms of the people who volunteered for the study, just simply based on the nature of that study. I myself, for example, could much more easily imagine myself wanting to be involved in, say, a Star Trek experiment or a science fiction experiment, than a prison experiment. So you can perhaps imagine how there may be some selection bias in terms of particularly cruel or sadistic

or whatever people joining up for that study. As I said, the other key finding of this experiment, apart from the power crops thing, is that it seems to lend support to a situational attribution of behaviour rather than a dispositional. Now, what that means is that the situational attribution of behavior means people behave the way they do because of the situation they are in. And if you put them in the right situation, they'll behave in a certain way. The dispositional attribution of behavior says that people behave the way they do mostly because of who they are, something internal to them, whether that be genetics or the way they were brought up or whatever. and they will behave in sort of similar ways

regardless of the situation they're in. Now, most people would say there's elements of both, but the question is how important is situational versus dispositional elements in people's behaviour? And the Stanford Prison Experiment leans very strongly in the situational attribution of behavior model. It seems to be the situation people were placed in that's most important rather than the dispositions of those individuals. Because ordinary, apparently, psychologically stable individuals, placed in this unusual environment and we seem to get the sort of behaviours that we would have

expected from much less stable, much less normal individuals. And I think that's perhaps the key takeaway message from these three studies is that situation matters a lot for human behavior. The Milgram Obedians experiment, I think, is the best illustration of that. And so we'll talk about that later on the episode. Now, as I said, there have been a number of criticisms of the Stanford Prison Experiment based on its methodology, and so some of the findings have been disputed on those grounds. And partly in order to address some of those criticisms, there have been a number of replications or, you might say, quasi-replications of the Stanford Prison Experiment. One of them recently was what you might call the BBC Prison Study, which was conducted in Britain in 2002

by the BBC. They obtained very different results with a fairly similar setup as Zimbardo had. It wasn't conducted at a university, but I think it was a purpose, a custom-built facility that was like a modern prison, and you had guards and prisoners and all that sort of thing. In this study, rather than the guards behaving sadistically and authoritatively towards the prisoners, the guards actually seemed reluctant to exercise their authority, to the extent that the prisoners actually launched a rebellion and overthrew the guards and sort of took control of the prison themselves. This may have been owing to a lack of clear leadership amongst the guards or a clear mandate to be able to use force to enforce, physical force to be able to enforce prisoner compliance.

So quite interesting that a similar setup had such different results. From my reading about that study, it seemed like from the beginning, the guards were never very comfortable with their situation or with their position and were never very willing to enforce the rules properly or use their authority. That may have been, because they weren't very many, I think they were only like four of them or something, maybe six. Similarly in the original study, that may have been simply due to a random, difference in the characteristics of the individuals being more or less prone to be willing to use the force or enforce the rules or whatever. It's hard to say because we don't really have the data

on that, but it is interesting. And if you want more information on that, look up BBC Prison Study, there was a television series made about it. Another interesting case which, perhaps not as well documented as one might like, is called the Third Wave. It's worth looking up. This was an experiment conducted in an American high school in the 1960s. Now, once again, It's hard to find out exactly how well-documented this is, but it does seem like that it really happened. Basically, from what I've read, a history teacher attending, a history teacher who had a contemporary world history class, they were doing a study of Nazi Germany. So as part of this class, the teacher was trying to explain to his students how the German population could claim ignorance of the Holocaust

and how they were so willing, not seemingly willing to support Nazis and all that sort of thing. So he was sort of having trouble convincing his students of this and explaining it to them. So he decided to show them instead. What he did was he sort of built a movement in his class and with other students as well apparently called the third wave, which he told his students was aimed at eliminating democracy and they had their own salute or something like that and their own slogans. And they developed an in-group, out-group sort of feeling. Once again, it's worth reading up a bit more about this to get a feel for what was done. But it didn't exactly, I mean, it wasn't a replication of the Stanford President experiment,

but what it did is would seem to indicate the presence of similar dynamics going on about people being very sort of malleable to the situation therein, that if they're placed in a situation where strong group identity and conformity is valued, then that is something that you will tend to see reflected in their behavior. They will behave in accordance with the environment they're in. once again a situational rather than dispositional explanation of behavior. The final experiment that we're going to look at in this episode is the Milgram Obedience experiment, and this is by far in my view the most disturbing of the three. This is not quite as famous as a Stanford prison experiment, perhaps, but still rather well-known.

Okay, so in the original experiment, Milgram, who was a psychologist, I believe the original was conducted in the 60s at some point. The original experiment was characterized as a study about the role of punishments in helping learning. That was what the subject were told, although that wasn't the real purpose of the experiment. But the setting was that the subject was given the role of being a teacher and a confederate of the experimenter, although the subject didn't know they were a confederate, was given the role of being a learner. The teacher and the learner were placed in two separate rooms where they could communicate, I think by a radio or telephone or something like that,

but they couldn't see each other in the two different rooms. And the experimenter was placed in the same room as the teacher, and they were instructing the teacher what to do. So the subject or the teacher was told that the... was told, as I said, that the purpose of the experiment was to study the effect of punishment on memory and learning. The teacher was placed in front of a large electric shock generator, which had 30 switches on it, each labelled with a voltage.

going from 10 to 450 volts. And at the switches at the high end of the voltage range were they had labels like extreme shock and danger severe shock, and the very last one had triple X at the end, so to indicate danger and potential for harm. The subject was given a sample of 45-volt shock to convince them that the apparatus was real, that it really did deliver electric shocks.

So this is an electric shock generator, which looked big and scary, basically, and it had lots of different notches on it, corresponding to different volts. The teacher was then given a list of word pairs that he was to recite to the learner and the learner was to recite back to them in the correct order or the correct pairs or something like that.

If the learner got it right, then they moved on to the next pair. If the learner got it wrong, then the teacher was to administer a shock to them. And the voltage of that shock increased in 15-volt increments for each wrong answer. So the more they got it wrong, the higher voltage shock they received, or the learner received. The learner, by the way, just press the button to indicate which of the options they selected for their response.

So the teacher can see which of the options that they selected and tell if it was the right one or not. The learner was led to a... So once again, we said that the learner and the teacher are in different rooms. The teacher saw while the learner was led to the other room and was strapped into what looked like an electric chair. So the teacher was shown this. Remember, the teacher is the real subject. The experimenter applied electrode paste to the learner's arms to, as they said, assure good electrical contact and strap the learners' arms into the chair so that it didn't look like they could escape. So the teacher was shown all this sort of to convince them that the learner was really in the chair and they were really receiving shocks.

And as I said, they also had the, they also received the sample 45 volt shock to convince them the device was real. So the key point is that the real subjects, the teachers, believed that the learners were receiving real shocks for every wrong. answer. In reality, this wasn't the case. In reality, what was happening is that there were no real shocks administered, except for the 45 sample one, that that was real. But the learner was a Confederate who was just acting, going through the motions. But the key thing is that the subjects thought that it was real. And as far as I could tell, there's never really any evidence of people suspecting otherwise. Not for the most part anyway. After the Confederate was separated from the subject, the Confederate set up a tape recorder, which was integrated with the shock generator, so that it played

pre-recorded sounds for each shock level. So the same sounds were played at a given shock level for each participant in the experiment. So that was consistent throughout. And these pre-recorded sounds went more or less as follows. At 150 volts, the learner began to complain loudly about a heart problem and demanded to be let out of the experiment. At 270 volts, they began to scream when shocked. At 330 volts, after a very intense scream when repeatedly demanded to be let out, the learner fell silent and made no further responses for the remainder of the shocks. And I believe that near the end of it, they stopped, the learner stopped responding as well, so they stopped giving answers.

And the teacher was instructed to just continue delivering shocks then to treat a non-answer as a wrong answer. I believe there were also other things like banging and yelling as well and other complaints at various points. But yeah, that's the key idea. So screams, banging, complaints, wrong answers, and no answers getting more severe as the number of the voltage of the shocks increased. Now, if at any time during the experiment, the subject indicated their desire to stop or not to deliver any more shocks, remember this is the teacher, who's the actual subject, they were given a succession of verbal prods by the experimenter in this order. Same every time. The first was, please continue. The second was, the experiment requires

that you continue. The third was, it is absolutely essential that you continue. And the fourth was, you have no other choice, you must go on. Now, if the subject still wished to go on with the experiment after four successive verbal prods, the experiment was halted. So if the subject was really insistent, after all those four prods, they did actually stop. Otherwise, the experiment was only halted after the subject, the teacher, had given the maximum 450 volt shock three times in succession. There were also specific prods, which were written up and given, which were decided on beforehand, and then given by the experimenter, if they were, the teacher made specific comments.

So one thing that a lot of people asked was if there was any permanent physical harm that was caused by the shocks. The experimenter to that was always replied, the shocks may be painful, but there is no permanent tissue damage, so please continue. Another thing was that the experimenter was very clear in that they would take all responsibility for any problems or any results of the experiment that the teacher was not to worry about that. Okay, so to recap, we've got a teacher delivering what they think to be electric, shocks to a learner who is actually not receiving shocks, but the teacher thinks they are.

The shocks being of successively higher voltages as the experiment continues, and with the learner in apparent distress and great pain, and potentially actually having suffered a heart attack or died near the end of the experiment because they stopped responding and the voltage was quite high. So what do you think would be the result of this? Do you think, remember, that all you had to do in order to get the experiment to stop as a teacher was to refuse to administer the shocks, even after four successive verbal prods,

which were actually quite subtle, not subtle, which were actually not particularly forceful, if you think about it, please continue the experiment requires that you continue, it is absolutely essential that you continue, you have no other choice, please, you must go on. That's not exactly yelling at someone or really trying to persuade them. So you might think that if people thought the experiment was wrong

or was unnecessary or whatsoever, that they would just ignore such prods and, and stop anyway. Before conducting the experiment, Milgram asked a number of Yale University psychology majors to predict the behavior of the teacher, so predict the outcome. I think he also asked some professional psychiatrists or something like that about the results as well, before the results were actually, before the experiment was conducted. Only a very small fraction of the teachers, something like 1%, were predicted to be prepared to inflict the maximum 450 voltage.

So people didn't think it would be very common. Basically, they just thought a few whack-job psychopaths would be willing to do that. Anyone else would not administer such high shocks to an innocent person for no reason just because some guy tells them to. Some, well, apparently legitimate authority, as the experimenter was. The experiment was wearing a white lab coat, by the way, and it was conducted at the university, which is important, and we'll talk about that a bit more later on.

So there was an element of legitimacy there. Okay, so what percentage of participants do you think continue to administer shocks, even in the higher range. As I said, at the time, psychology majors and psychiatrists thought it'd be about 1% would go on to the 450 final shock. As you might imagine, it was higher than that. The answer was 65%, 26 out of 40 participants

administered the experiment's final 450-volt shock, though many were very visibly and verbally uncomfortable in doing so, but fully two-thirds of everyone in the experiment delivered that final shock. And fully 85% of the participants continue to administer strong and very strong shocks, even after the learner had expressed substantial distressed and asked to be let out of the experiment. So an additional 20% of people were willing to administer strong,

very strong shocks to someone who was clearly under distress and had specifically asked to be let out of the experiment. So that's a very large majority. And at some point, every single participant, all of the thought he asked, so stop, stopped and questioned the purpose of the experiment, many were visibly uncomfortable administering the shocks. Some said they would refund the money they were paid for participating. Many displayed nervous laughter, stuttering, swearing and trembling,

repeatedly asked if they should continue, asked if the man was okay, asked who would take responsibility. As we saw in the Ashken Formity experiment, people were very, very uncomfortable doing this, but almost everyone did. Or I should say, 65% of people went all the way, and 85% of people went a good portion of the way. So that's still 15% of people who refused to administer even strong shocks.

There is a graph which I saw somewhere which shows exactly how the behaviour of the teachers dropped off as the experiment continued. Interestingly, I think virtually I think 100% of the subjects were happy to administer the shocks initially and for a while

even though they were painful. No one dropped out until the, if I'm remembering correctly, until the learner requested to be left to be allowed to live experiment and started to scream and really make it make an issue. But even there only like 15% of people dropped out around that point. Most people continued going on further and mostly to the end. Okay, so as with the Stanford Prison Experiment, you might imagine that this

is just some weird result. I mean, it's not a very large sample size. It's only 40. Perhaps there was some selection bias there. Perhaps Milgram had some, had designed the study incorrectly, or there was something else strange going on. So this experiment, much more so than the Ash conformity or Stanford Prison. This experiment has been replicated many, many times in dozens of occasions. Many by Milgram himself, but many also by later experimenters. And as with the Ash conformity experiment as well, there have been various manipulations and alterations of the conditions of the experiment to see how they would change the results. But the general finding that most people can form is very robust. So Thomas Blass, Blass, performed a meta-analysis,

which is an analysis of experimental results, basically, of a large number of them. And you pull them together and try and find the overall effect. Anyway, he performed a meta-analysis on the result of repeated performances of the Milgram experiment. He found that the percentage of participants who were prepared to inflict fatal voltages, so fatal shock voltages, on learners, remains remarkably constant, around 60 to 66% of participants, regardless of time or place. So Milgram's results were quite typical, they weren't a fluke. This has been replicated on different people in different environments, in different circumstances.

They seem to be quite robust. So that's quite disturbing. Let's look at some of the more specific variations and see how they could increase or decrease the percentage of people willing to administer shocks. So as you might expect, one obvious thing to change is the immediacy of the teacher and the learner. Generally, when the victim, so the person being shocked,

that is the learner, when their physical immediacy was increased, that is when they were moved closer to the subject, when the subject could see them and the closer they were to them, the more, the lower was compliance. So people are less willing to shock people that they can see and are in direct contact with compared to people in another room. The reverse effect is observed with the physical immediacy of the experimenter. The closer the experimenter is and the more directly they are in contact with the teacher, the more the higher was compliance. In one of the variations of the experiment where participants received instructions from the experimenter by telephone,

as opposed to the experimenter being in the room with them telling what to do. Compliance all the way through decreased to 21% down from 65%. Very interestingly, a number of participants pretended to continue with the experiment by administering the new shocks in accordance with the experimenter's orders, even though they weren't actually doing it. So they were faking obedience to the experimenter, even though they knew it was an experiment and they could have just said, I don't want to do this. But apparently it's socially difficult to define it.

authority in that way and we'll come back to that when we do the analysis of this experiment. One variation of the experiment was held where, this is the one where the physical immediacy of the learner was greatest, the participants actually had to physically hold the learner's arm onto a shock plate, although it wasn't a real shock plate, but they thought it was. Only 30% of subjects went to the end of the experiment and delivered the 450 volts shock. But still, that's about one in three people willing to deliver fatal voltage shocks to people while physically holding their arm to the shock plate. That's quite surprising.

In another variation, women were the participants, I believe it was only men in the first version, many of the versions. Obedience was not significantly different. Another variation of the experiment conducted by Milgram was not held at the university. As I said before, one proposition was that being held at and having a connection to Yale University lent an air or a significant feeling of legitimacy towards what was going on, which was affecting the behaviour. So in this version, it was held in a rather dingy office with no apparent connection to Yale University.

In these conditions, obedience dropped to about 48%, although the sample size was, I think, fairly small in this condition, but still relatively high obedience. Another thing he did was sort of combine the Ash Conformity experiment with his own obedience paradigm by incorporating additional Confederate teachers. So these are additional teachers in the experiment, people delivering shocks, except they're not really teachers. They're not really subjects. They're confederates of the experimenter. But the real subject, the real teacher doesn't know that. Now, remember, in the Ash Conformity experiment, we found that the existence of only one dissenter was enough to basically eliminate the effect of conformity. And we find a very similar thing in the Milgramobinance experiment.

In one version where two additional teachers refused to comply with delivering shocks, only 10. 10% of participants continued on with the experiment, which is perhaps still more than you would have thought, but a lot less than 65. Similarly, there was another version of the experiment where you had Confederate teachers who, instead of refusing to comply, did comply with the experimenter. Almost everyone continued with the experiment. So the presence of other people obeying increased compliance, which is again not surprising given the ash conformity results. One of the biggest things you can do, so remember, you're holding in a dingy office, obedience goes down like 20% or something like that. You have women as the subjects

who you might stereotypally think to be more empathetic or something doesn't really make any difference. You have to physically hold someone's arm to the shock plate. Abedance goes down by about half, but it's still at 30%. By far, one of the biggest things you can do

to reduce compliance is to replace the original experimenter by another person who was previously identified as basically being a clerk or a secretary or something like that. In this version, only 20% of subjects obeyed fully.

So this points to a key aspect of the reason for the obedience here is authority. And this is really what it seems that the key result of this experiment is. It's the apparent legitimate authority of the experimenter in the fact that they were sort of conducting the thing. They were in charge. They were wearing a white lab coat. I don't know if they were wearing a white lab coat in all variations, but certainly in the initial ones they were. I've seen photos of that. So they appeared legitimate. They appeared like they knew what they were doing and they were willing to take responsibility. When they were absent or distant or replaced with some

cleric, obedience dropped down dramatically. Now, these ones that I've talked about, these specific results are mostly Milgram's replications, but many exact and variations of the experiment have been conducted in different countries with different types of participants, so not just college students, and as I said, the meta-analysis shows very robust results. So, I mean, that leads to the question, is this a lot of conformity or obedience, or not very much? Well, it's clearly much higher than most people thought at the time, you know, like 60% compared to 1%. However, on the other hand, that's still, you know, one in three people who at some

point refuse to follow orders, and in some of the variations, most people refuse to go through with it. So there is perhaps some cause for optimism there. And there's one person who argued here that, oh, let me just read his quote. People have learned that when experts tell them something is all right, it probably is, even if it doesn't seem so. In fact, it is worth noting that in this case, the experimenter was indeed correct. It was all right to continue giving the shocks, end quote. So that was Robert Schiller there, who's a professor of finance at Yale. It's a very good point, actually, that people generally trust experts, especially if they're associated with science, in a given field. And so if the expert tells you it's okay, you have a general

heuristic, that is sort of a rule of thumb, that it is okay, even if it doesn't seem like it's okay. And it turns out it's actually valid in this situation because it was okay. The person wasn't really receiving shocks. It may not be so much an issue of obedience so much as an issue of people were getting information from the fact that an expert was telling them to do something that it must have been okay, an apparently legitimate expert nonetheless, which was similar to what we saw in the Ash Conformity experiment, whereby it's perhaps arguably not so much an issue of conforming to the social norm, but an issue of getting information from what other people were doing. So the findings of the Milgram-Obenians experiment have been applied to a number of real-world situations

to explain, for example, the Holocaust or another wartime massacres, the Ma'a-Lai Massacre is another example in the Vietnam War, where basically, you can imagine in something like a concentration camp, if you have legitimate authorities, which would be, for example, higher-ranking members of the SS or military commanders or something like that, if you have other people conforming or obeying, and at least some people are going to do that just because they're sadistic, essentially, even if that's a minority, at least some people will do that, and that's, as we saw, is going to increase obedience. If you have a high level of immediacy of those superiors to you when you're carrying out your orders,

as you might, for example, at a concentration camp or a military facility, and if you have a fair degree of distance between you and those whom your actions are affecting. So, for example, if you're just pulling the levers to release the gases in the gas chamber, and you don't actually have to see what's going on in there or actually see the people that you're killing. All of those things are going to increase obedience as we see from the experimental results. And so it's not as surprising, perhaps,

that you can get high levels of conformity and obedience in those situations, even though most people, most of the people in that environment might have, you know, if you'd ask them beforehand, have said that they would never do such a thing. Because I think it's one of the things that is most disturbing about these experiments that we've talked about is that they, as I've mentioned, they point heavily towards the situation rather than the dispositional theories of

behaviour. Now, most people don't like to think in those terms. They like to think that they are a good person and other people do bad things because they're bad people. And there's likely some truths to that, of course, but these experiments all point towards, well, not so much. You and most people, well, no one really, but most people in particular are not immune from these effects and that if you were placed in this situation, you probably would obey. In the highest versions of the experiment, like 85% of people or 90% of people obeyed. Yeah, 90% I think was about the highest obedience that was recorded with the right set of circumstances.

So about 90% of people, it seems, are willing to kill other people or, you know, probably kill them or at least harm them, given the right situation and legitimate authority-giving orders and so on. And there's actually a psychological theory which explains why that's the case. this is known as the egentic state theory. And let me read a quote here. The essence of obedience consists in the fact that a person comes to view themselves as an instrument for carrying out another person's wishes, and they therefore no longer see themselves as responsible

for their own actions. Once this critical shift of viewpoint has occurred in the person, all of the essential features of obedience follow. End quote. I believe that's from Milgram himself who said that. So that seems to be the key thing that was occurring here. People seeing themselves as an instrument of the experimenter. And this was reinforced by such things as the experimenter saying that they would take responsibility for anything that happened by the experimenter, demonstrating their authority and demonstrating that they were in charge and demonstrating that they said it was okay, so they knew what was going on. The person was just an instrument in their hands, and when people come to view themselves as that way, they feel disassociated from their own actions, and they're more willing to go along with it.

There's one very interesting real-world version of this experiment which was conducted. It's called the Hoffling Hospital Experiment from 1966. a psychiatrist called Hoffling conducted a field experiment on obedience in the nurse physician relationship. So in a natural hospital setting, nurses were ordered by unknown doctors to administer what could have been a dangerous dose of a fictional drug to their patients. So the drug wasn't real, but it could have been, it was a dangerously high dose of whatever it was, that the doctors were ordering be administered by the nurses. Now, there were official guidelines which forbade the administration in such circumstances. I don't even think that the nurses were supposed to administer, was supposed to administer,

to conduct orders from unknown doctors without like extra signed orders or something like that. And they weren't supposed to give dangerous dose levels if they thought it was, or if it was beyond a certain safety limit or something like that. I don't know exactly what safety protocols they had in place, but they weren't supposed to. But in spite of these official guidelines, 21 of the 22 nurses would have given the patient the overdose of the medicine. I don't think, I think they were stopped before they actually gave it, but they were going to do it.

That's basically all the nurses were going to ignore the official guidelines and follow the orders of a doctor, basically just because doctors have an authoritative position, it seems. And perhaps the most bizarre thing about this is that before the nurses did the study, he actually asked them if they would obey the orders of the doctor in this circumstance. And all of them said they wouldn't obey orders. And then basically all of them did obey orders. So that seems to point towards a previous,

episode that I've done about the introspection illusion, episode seven, the introspection illusion, where people are horribly bad at working out their own motivations and the reasons why they do things. And that, once again, this would also point towards the situational and away from the dispositional theories of behaviour, because when the nurses were asked, would you do this? They were thinking, dispositionally, would I do this? Well, no, I'm a careful person. I'm a caring person. I would follow the regulations to be safe. I wouldn't just follow the orders from some strange doctor I didn't know that apparently told me to do something that was unsafe. But when they were actually in the situation, they found themselves acting very differently. And they didn't expect that

when they were answering the questionnaire because they weren't considering that situational element of what was going on. The scariest part of that study was that 10 of the 22 nurses had done this before with a different drug, had given an apparently overdose or something similar based on the orders of a doctor. So this was just one little study in one hospital in the 60s. You can imagine how often this sort of thing happens in real life in hospitals and in other circumstances as well, where people do things that they know are likely to be wrong or stupid or something like that, but just follow waters anyway for these sorts of reasons. Okay, so on that very depressing note, that's the end of this episode.

We've covered the Ash Conformity, the Stanford Prison and the Milgram-Mobinance experiments. But I hope you enjoyed this episode. If you did, please tell other people about the podcast. I need more listeners. I also need more ratings on iTunes. So log on there and write me a review, give me some stars. Thanks for listening and I'll talk to you next time.