Demystifying Psychology

Thursday, March 15, 2012

Milgram's 'Shocking' Experiments on Obedience to Authority

It would be difficult to discuss ethics in psychological research without introducing the infamous experiments conducted by Milgram in the 1960's.  

Following World War II, Milgram became very interested in what led people to commit such atrocious acts as those witnessed in Nazi Germany. He aptly hypothesized that the presence of authority (or perceived authority) played a key role. It is my understanding that this hypothesis may have been strongly influenced by the results of the Nuremberg trials. During these trails, many men claimed that they were good people who were driven to do terrible things by the orders of Nazi authority figures.

About the Experiments
Milgram set out to test obedience to authority in a series of laboratory experiments.
In the first experiment (and perhaps the most well known), adult working men were invited to participate in a "study of memory and learning"at Yale for $4 in compensation. Here is a copy of the study advertisement, which I find to be particularly interesting:

Milgram (1965) states that he selected a "balanced group" of participants by evenly representing different occupations and age groups. All of the 40 participants in this first experiment were male.
When participants arrived at the laboratory, they met another man who identified himself as a fellow participant, but was actually a confederate actor. The two men were given what Milgram terms a "general talk on the effects of learning and memory". Then, a rigged "drawing" was held in which the confederate was always given the position of the "learner" and the true participant was always given the position of "teacher". The confederate "learner" was then taken to another room and strapped to what is termed an "electric chair". The participant "teacher" was led to an adjacent room and instructed to read the "learner" a set of word pairs to memorize. The "teacher" was told he must administer a shock each time the "learner" fails to respond with the correct word, and that the shocks must increase in intensity each time the "learner" gives an incorrect answer (the participant was even given a "sample" shock of 45 volts to convince him of the "authenticity of the instrument"!)

The confederate "learner" being strapped in
The naive participant "teacher" being given instructions
The confederate "learner's" responses were recorded on a tape and played back in response to the participant's actions, and the recording included frequent wrong answers so that the participant would need to increase the level of shock. Milgram was really studying how far the participants could be made to take this shock punishment, represented by a board of switches in increments of 15 volts. The switch board began at 45 volts (labeled "slight shock") and continued all the way to 450 volts (labeled "Danger: severe shock"). In the original 1930 experiment, Milgram even added two additional switches beyond 450 volts, labeled simply with "XXX" (which I still find a little amusing to this day...but that amusement is dampened somewhat by the sobering results of this experiment). For those of you who are about as skilled in mental math as I am: that is 30 switches in total!

When the participant flipped the 75 volt switch, the "learner" began to grunt. At 150 volts, he said he wanted out of the experiment. At 180, he said he couldn't "stand the pain" any longer. Finally, at 300 volts he refused to answer any more memory questions and said he wanted to be freed. In response, the experimenter told the naive participant that he should treat no response as an incorrect response and proceed with increasing shock voltages. According to the procedure, the study would end if that participant ceased giving shocks to the "learner". But here' s the catch. If the participant "indicated his unwillingness to continue", the experimenter was instructed to use the following "prompts": "Please continue","the experiment requires that you continue", "It is absolutely essential that you continue", and finally "You have no other choice, you must continue!"

And continue they did. Shockingly (no pun intended...) all 40 participants proceeded to 300 volts, which was actually well beyond the predicted stopping point when Milgram had polled senior psychology majors and psychiatrists alike. What's more, 25 participants then continued to shock the now-silent "learner" all the way to the end of the board, where the switches were labeled "XXX" (see? not so funny anymore...). Bear in mind that just because these participants continued with the experiment for so long, does not mean that they were not greatly upset by what was transpiring. I tried to find the best video I could which shows some of the participants' reactions to what they were doing. Although the video above only shows tiny glimpses, many participants covered their face with their hands in distress, fidgeted, smoked, rubbed their eyes, shook their heads, or even laughed (you know...that eery laugh when you are overwhelmingly upset?). Some even got angry. Many of them were clearly experiencing painfully high levels of anxiety over what they believed they were doing to the man on the other side of the wall.

After the experiment had ended (either when they had used every shock voltage or had refused to continue), the participants were allowed to have what is termed a "friendly encounter" with the confederate who had been playing the "learner". They were informed that they had not really been shocking him at all, and obviously he was OK. This was followed by a brief interview which appears to have been an attempt at debriefing and ensuring participants were psychologically healthy before leaving the laboratory.

Although some of the other variations of Milgram's first experiment are not cited as often, I would like to at least briefly touch on them, as I feel they are extremely relevant to this discussion.

In the next experiment, he investigated the effect of the proximity, or immediacy, of the "learner" to the participant. He used several degrees of separation, including: a wall with no sound other than the "learner" kicking the wall at 300 volts; a wall with verbalizations and sounds like the first experiment; a visible and audible "learner" with no wall but no physical contact; and finally no wall with physical contact ("learner's" hand actually had to be held down by the participant). As you might predict, with each growing degree of immediacy the participant's obedience to authority significantly decreased. To learn more, take a look at his article.

He also designed an experiment to examine the closeness of the authority figure. The "experimenter" was either a few feet away, outside the lab giving instructions via phone, gave instructions and left the room, or was merely a tape-recorded voice giving instructions. Also as you might guess, the more 'distant' the authority figure was, the less likely the participant was to obey. However, if the experimenter returned to the room after the participant refused to follow through, he could usually coerce them to continue as in the first experiment.

The effect of group pressure, both good and bad, was also examined. As you might expect, when two other confederates in the room decide not to obey, 90% of the other participants follow suit, presumably because they wanted to as well and this social validation helped them act on it. Furthermore, when two confederates in the room immediately obey the authority of the experimenter without question, this increases his authority "only slightly", also indicating the participants weren't sadistic (in my opinion).

Finally,when Milgram moved the location of the laboratory to a local building downtown and removed the prestige associated with Yale University, he did see a drop in obedience among participants. However, this drop was not significantly different from the level of obedience found in the original experiment.
For more information on any of these studies, please read Milgram's article summarizing them.
Experiments on women and different types of authority or conflicting authority were also carried out, and are detailed in Milgram's book.

Analyzing the Experiments
One of the major strengths of Milgram's work is the fact that he actually conducted a series of experiments, many of which examined potential third variables and situational characteristics which might impact the research. He used different locales, different set-ups, different group sizes, and even examined gender differences. However, Milgram was still a human experimenter and some biases are obvious. For example, the participants were "selected". It is always better to use random sampling whenever possible to avoid the bias that surely affected some of Milgram's selections. It is possible that some third variable common to all the participants who made it through the selection process might have affected the results of the study.
The many situational factors which Milgram did examine, however, did seem to have an impact on the results. They all pointed to obedience to authority as a powerful force, but suggested that certain conditions can mediate it. I personally find that to be a rather encouraging part of the research, though it is often neglected in favor of the more shocking results of the initial experiment.
Furthermore, I find the common interpretations of Milgram's results to often be extrapolated a little to far. There is a rather large difference in my mind between tourturing and killing a fellow human standing in front of you --especially an unarmed and innocent civilian-- and pressing a button which gives a man who cannot be seen an electric shock as part of an experiment. Furthermore, the authority of the experimenter is significantly reduced in effect in any situation which provides validation in the form of a fellow participant, or puts the victim closer or the authority figure farther away from the participant, as is evidenced by many of the follow-up experiments.

An Ethical Eye-Opener
Obviously, one of the biggest controversies surrounding Milgram's infamous experiments is the ethical implications.

First, while the participants were told at the beginning of the study that they were free to leave at any time, the use of "prompts" in an attempt to coerce them to stay would not fly by today's ethical standards. Once the participant says he or she wants to stop, they need to be allowed to stop. Rather, the participants in Milgram's experiment had to physically refuse to flip the next switch in order to end the experiment. However, it is important to remember that this experiment would not have validly tested for obedience to authortiy without such coercion.

The next obvious ethical area is pain and discomfort (which can be emotional and psychological just as much as physical). As you can see from the following chart of self-report measures, the participants experienced very significant level of discomfort during the course of this study.

And this discomfort was mirrored in their behaviors and reactions. As one observer wrote:
I observed a mature and initially poised business- man enter the laboratory smiling and confident. Within 20 minutes he was reduced to a twitching, stuttering wreck, who was rapidly approaching a point of nervous collapse. He constantly pulled on his earlobe, and twisted his hands. At one point he pushed his fist into his forehead and muttered: "Oh God, let's stop it."  (Milgram, 1963).
Even Milgram himself seemed disturbed at some of the participants' nervous laughter.
 The effect of this experiment on its participants can hardly be denied. What's more, the participants were not at all informed of what they would undergo in the study, and were severely deceived. While I am a big fan of ethical deception and do believe it has a very important place in psychological research, guidelines today state that it should not even have a moderate psychological effect on the participant and that measures should be taken to avoid any longterm damage. While Milgram did (sort of) debrief his participants, it was rather perfunctory and I would argue certainly not sufficient for the amount of stress the study entailed. He also failed to sufficiently follow up with his participants after they left the experiment. This is not to say that there was no follow-up, but that it was not at the level of psychological therapy that really would have been necessary for such a taxing experiment. In reading about it (see this article), it seems that Milgram was more interested in medically proving that he had not harmed his participants than in providing them with resources for emotional healing and healthy self-acceptance. 

Milgram himself wrote a response to many of these criticisms that sheds light on some very interesting and perhaps peculiar participant reactions to his studies.
As you can see in the table above (from his response article published in the The American Psychologist), nearly half of participants reported feeling "very glad to have been in the experiment", and only 16% were "neither glad nor sorry", "sorry", or "very sorry" to have participated.
This seems to me to be an attempt for the participants (as well as Milgram himself, perhaps?) to make themselves feel better about something that cannot be undone. But this is definitely all my personal opinion. It might really be possible that these participants truly felt they learned something about themselves (as they also reported in this follow-up survey). I just tend to believe that what they learned about themselves was extremely depressing and psychologically damaging, making this line of questioning invalid.

Interestingly, Milgram later authored an article in 1977 on subject reaction, and described it as being an often neglected aspect of ethics in psychological research. It certainly would be satisfying to know that he learned from his initial mistakes. But even if Milgram himself didn't, the rest of the psychological world most certainly did. The concept of Internal Review Boards followed closely on the heels of this momentous study.

Follow-up Research and Meta-Analyses
In 1977, Burley and colleagues found that a measure of individual difference also contributed to obedience. They tested participants' social intelligence levels before administering Milgram's experiment, and found that social intelligence did indeed affect obedience to authority.

More recently, in 2008, Packer performed a meta-analysis of Milgram's studies and found that the 150 volt mark (the point at which the "learner" first expresses pain and asks to be freed) was actually the turning point for the experiment: at that point it is suggested that participants made the decision to continue administering shocks, or to stop. If you are interested in reading more about this study, it can be accessed here.

In Conclusion...
Although this is one of those studies that doesn't exactly make us feel rosy about what we are capable of, it is a fascinating experiment nonetheless. Perhaps that is because there is still a part of us that refuses to believe that we could ever be persuaded to act as nearly all of Milgram's participants did. Don't kid yourself: these people weren't "crazy" and they don't belong in prison (which is a comment I frequently hear from people who aren't very familiar with Psychology). You very well could have been right there with them. Obedience to authority is a powerful force, and so is group pressure (which will most certainly be addressed later on this site). It's how we're wired. However, our social wiring is something that can be used for great good just as easily as great destruction, and that is what is important to remember.

Read about it! Talk about it!
The original article which Milgram published about his first in 1963 can be found at this location.
Another article which discusses other versions of the experiments published in 1965 can be read here.
I would like to encourage you to explore this topic for yourself and add to this discussion. What do YOU think?

Burley, Peter M.;McGuinness, John (1977). Effects of social intelligence on the Milgram paradigm. Psychological Reports, 40.
Packer, Dominic (2008). Identifying Systematic Disobedience in Milgram’s Obedience Experiments: A Meta-Analytic Review. Association for Psychological Science, 3: 301-304.
Milgram, Stanley (1964). Issues in the Study of Obedience: A Reply to Baumrind. American Psychologist, 19:848-852.
Milgram, Stanley (1963). Behavioral Study of Obedience. Journal of Abnormal and Social Psychology, 67: 371-378.
Milgram, Stanley (1977). Subject Reaction: The Neglected Factor in the Ethics of Experimentation. The Hastings Center Report. 7: 19-23.
Milgram, Stanley (1965). Some Conditions of Obedience and Disobedience to Authority. Human Relations. 
Photo credits:
Personal Collection of Alexandra Milgram, via NY Times 2008 
"The Situationist" blog located at  
Wikimedia Commons, public domain
YouTube video


  1. You can also learn more about modern day psychologist at: