Were Milgram's Obedience Studies Valid?

Let’s say that we wish to test the hypothesis that smoking causes lung cancer. In order to test this hypothesis, we consider the case of a man who smoked three packs of cigarettes per day for 70 years, and then died at the age of 90 in a car accident. Can we conclude from this case study[∂] that smoking does not cause lung cancer? No, we cannot: there may be extraneous variables[∂] that protected the man from developing lung cancer, such as "good" genes and an otherwise healthy lifestyle. If a correlational study[∂] of 10,000 people produces a positive correlation[∂] between smoking and lung cancer, can we conclude that smoking causes lung cancer? Again, no, we cannot: correlational studies are limited by the directionality[∂] and third-variable[∂] problems.

The conclusions that we are able to draw from the results of a study depend on its validity — its ability to help us understand the phenomenon under investigation in meaningful ways . A research study can be evaluated in terms of two main types of validity: internal validity and external validity.

Internal Validity
Internal validity refers to a study's ability to provide an adequate test of an hypothesis — its ability to rule out alternative explanations (other plausible hypotheses) by controlling adequately for the effects of extraneous variables (such as third variables) and for the directionality problem. For example, let's say that researchers conclude that a particular medication prevents the development of heart disease because they found that people who took the medication showed lower rates of heart disease than did those who never took the medication. This conclusion, however, can be made only if the study has high internal validity, which would require at least the following: (a) participants were randomly assigned[∂] to the experimental and control groups; and (b) participants did not know whether they were taking the medication or the placebo (see Table 1).

The most internally valid studies are experimental studies because they control for the directionality and third-variable problems, as well as for the effects of other extraneous variables. As an example of experimental research with high internal validity, let's look at a series of experimental studies performed by Joseph Goldberger during the early 1900’s (Chase, 1980; Kraut, undated; Stanovich, 2004). Goldberger wanted to find the cause of a disease called pellagra, the symptoms of which include gastrointestinal problems, mental disturbances, and skin eruptions. Many medical doctors theorized that the disease was caused by some sort of “germ” that, when passed from one person to another, gave rise to the symptoms of pellagra. Because the disease was more likely to occur among people living in unsanitary conditions (such as in homes with no plumbing), some theorized that the unknown microorganism was passed through bodily wastes and fluids. Goldberger, on the other hand, suspected that the disease was caused by an inadequate diet, and that the correlation with sanitation was due to a third variable — poverty — that caused people to have inadequate sanitation and to receive poor nutrition.

In order to evaluate the theory that pellagra is caused by poor nutrition, Goldberger tested the following hypothesis: eating a low-protein diet will cause the symptoms of pellagra. His experimental design looked like this:

Goldberger sampled prisoners from the Mississippi State Prison (today, it would be considered unethical to use prisoners for such a study) and split them into two groups: the Experimental Group received the low-protein diet and the Control Group received the normal diet. Within five months, many of the prisoners in the Experimental Group developed pellagra, whereas none of the prisoners in the Control Group developed the disease. Because of its high internal validity, Goldberger concluded from this study that pellagra is caused by poor nutrition.

Goldberger performed a second study designed to disprove the alternative theory, which claimed that a microorganism passed in bodily fluids causes pellagra. Goldberger hypothesized that, if the alternative theory is wrong, then people who consume the bodily fluids of pellagra patients will not develop pellagra. His experimental design was as follows:

In performing this study, Goldberger took secretions from the nose and throat of pellagra victims and placed them into his own mouth. He did not get sick. But, because the microorganism may not have been at high concentrations in these secretions, he performed the following manipulations:

he selected two patients — one with scaling sores and the other with diarrhea. He scraped the scales from the sores, mixed the scales with four cubic centimeters of urine from the same patients, added an equal amount of liquid feces, and rolled the mixture into little dough balls by the addition of four pinches of flour. The pills were taken voluntarily by him, by his assistants and by his wife. (Bronfenbrenner & Mahoney, 1975, p. 11; quoted in Stanovich, 2004, p. xx)

None of these participants developed pellagra after consuming the bodily fluids. The control condition never had to be performed because Goldberger and his associates never developed pellagra from the fluids of people who had the disease. If a microorganism were involved, then it seems likely that at least one or two of the subjects should have become ill. By comparing the results of Goldberger’s several studies, we can feel very confident that pellagra is caused by an inadequate diet and not by a microorganism. Because of the care he took in controlling his research situations, Goldberger’s investigations into the causes of pellagra definitely had high internal validity.

Did the Milgram study provide an adequate test of the hypothesis that people will obey an authority's commands to physically harm a stranger? In other words, did the design of Milgram's study allow him to make observations that provided strong support for this claim? It now should be obvious that Milgram’s study had high internal validity: By manipulating particular variables (such as the immediacy of the Learner) one at a time while holding extraneous variables constant, Milgram was able to show that several situational factors were the primary causes of differences in levels of obedience.

External Validity
External validity refers to a study's ability to produce results that can be generalized to other situations, especially natural ("real-life") situations. For example, if the researchers in the fictional heart-medication example described above found that the medication, under controlled conditions, prevented the development of heart disease in their participants, they then would want to generalize these findings — to state that the medication will prevent heart disease in the general population. If, on the other hand, the study design required the exclusion of potential participants, such as those who abused alcohol or other drugs, suffered from diabetes, weighed more than average for their height, and had never suffered from a mood or anxiety disorder, it would be difficult to generalize the study's results to the general population. These are common risk factors for heart disease and we could not conclude based on the study's results that the medication would help prevent heart disease in people with these problems. In other words, the study would have low external validity.

Studies with high internal validity often have low external validity because, in setting up a research situation that adequately controls for the effects of extraneous variables, it is necessary to create an artificial situation — a situation unlike the natural situations to which we would want to apply the results. In other words, the greater the amount of control, the more unnatural or artificial the research situation becomes. In fact, the research situation may become so unlike any natural situation that the study's results have no meaning outside of that specific situation. For example, let's say that researchers are testing the effectiveness of a new antidepressant medication. Because they want to make sure that the antidepressant effects of the medication are not masked by the effects of extraneous variables — such as drugs of abuse, other medications, and other mental disorders — they decide to exclude potential participants who are addicted to alcohol or other drugs, take particular medications, have anxiety disorders, or exhibit psychotic symptoms. Furthermore, because people with mild depression can show only a small response to the medication, they also are excluded. In this example, most people suffering from depression have been excluded from the final pool of participants (see Antidepressant Drug Trials...). Thus, the study's results can provide no information about how most depressed people would respond to the medication. The study would have high internal validity but low external validity.

High external validity depends on two factors:

• Participants. The participants must be a representative sample of the population of people to which one wants to generalize.
• Physical Realism. The physical setting must be similar with respect to relevant and important characteristics of the natural situation to which one wants to generalize.

Milgram (1974) reported that he had been concerned about the representativeness of his sample when he performed his obedience esperiments. When he performed his first studies, he used undergraduates from Yale University and found that about 60% fully obeyed the Experimenter. Another Yale professor:

immediately dismissed these findings as having no relevance to ‘ordinary’ people, asserting that Yale undergraduates are a highly aggressive, competitive bunch who step on each other’s necks on the slightest provocation. He assured me that when ‘ordinary’ people were tested, the results would be quite different. (p. 170)

In order to make the sample more representative of the general population, Milgram sampled both white and black men from a variety of occupational backgrounds: professionals, white-collar workers, blue-collar workers, and the unemployed. He found that these participants showed about the a similar level of obedience (65%) to that of the Yale undergraduates (60%). To further demonstrate the external validity of his studies, Milgram performed studies of other groups. For example, he found that 65% of women showed full obedience. Furthermore, he reported the results of obedience studies performed by different researchers in other countries (West Germany, Italy, South Africa, Australia, etc.) with participants from a broad range of backgrounds. In all of these studies, the majority of participants fully obeyed the experimenter in the standard condition. In fact, the percentages were higher in some studies: for example, 85% of the West-German participants were fully obedient. These results show that it is an empirical question (that is, a question for which one must make the relevant observations) whether or not a particular characteristic of a study reduces its external validity. It is very important to remember that we cannot dismiss experimental findings based only on our intuition that the participants "obviously" were not representative of the population.

What about the fact that Milgram’s studies took place in laboratory settings — settings that "obviously" are very different from most real-life settings and, hence, have little or no meaning outside of the laboratory? If the answer to this question is “yes,” then the results of controlled laboratory studies never would have enough external validity to allow us to apply their results to natural situations. Milgram (1974), however, stated that people making this criticism expose their serious misunderstanding of science. He argued that, by creating a research situation that strips away the complexities of the natural world, controlled laboratory studies allow researchers to discover the basic processes that must exist if a phenomenon is to occur. That is, by constructing a simplified research situation, scientists are better able to test their predictions about how the natural world operates. Creating an appropriately artificial situation is a strength, not a weakness, of laboratory research.

Stanovich (2004) argued a similar point when he stated that, contrary to common belief,

the artificiality of scientific experiments is not an accidental oversight. It is intentionally sought. Scientists deliberately set up conditions that are unlike those that occur naturally because this is the only way to separate many inherently correlated [that is, confounded(∂)] variables that determine events in the world. (p. 99)

When designing controlled laboratory studies, researchers attempt to recreate particular aspects of the natural situation thought to be important for causing the phenomena being studied; but they do not want to recreate the entire natural situation. In the laboratory, researchers are trying to test a theory about one or more causal relationships. A contrived and artificial laboratory situation is more likely to reveal these causal relationships because confounds among variables that occur in natural situations are excluded.

Milgram (1974) argued that some basic processes underlying obedience to authority were revealed by his experimental studies, and that these discoveries were applicable to a wide range of natural situations. In fact, he concluded that the obedience shown by his participants to the cruel commands of the Experimenter was the result of “the capacity for man to abandon his humanity, indeed, the inevitability that he does so, as he merges his unique personality into larger institutional structures” (p. 188). This interpretations of his results led to his claim that this supposed capacity to abandon our sympathy and empathy for others “is a fatal flaw that nature has designed into us, and which in the long run gives our species only a modest chance of survival” (p. 188). There seems little doubt that his laboratory experiments revealed something important about human nature, although his conclusion that we have only a "modest chance of survival" because of a "fatal flaw" in our nature seems to be highly speculative and not justified by the results of his research. Nevertheless, this doesn't mean that the claim is incorrect, simply that he has overgeneralized[∂] his results.

Study Questions

1. How would you define, in your own words, the meaning of "internal validity"?
2. Would random assignment to different groups increase, decrease, or have no effect on the internal validity of a study? Why?
3. How would you define, in your own words, the meaning of "external validity"?
4. Would random assignment to different groups increase, decrease, or have no effect on the external validity of a study? Why?
5. Controlling for the directionality and third-variable problems found in correlational studies would increase which type of validity? Why?
6. What was Joseph Goldberger's theory of pellagra?
7. Goldberger was more concerned with maximizing which kind of validity in his studies? Why?
8. How did Goldberger show that pellagra was much more likely to be caused by inadequate nutrition than by an organism transmitted through bodily fluids?
9. Why would we conclude that Milgram's studies had high internal validity?
10. Let's say that we want to show that a new medication reduces high levels of cholesterol in people over 40 by performing a study of 1000 men over 40, 500 of whom receive the medication and 500 of whom receive a placebo. The study is performed in a double-blind manner. Which type of validity would be lower in this study: internal or external? Why?
11. Variations in which two factors determine the degree to which a study is externally valid?
12. Which factor was Milgram most concerned with when he tested people from the surrounding community in addition to undergraduate students?
13. Why are artificial research conditions created in laboratory settings if this procedure reduces the external validity of studies?
14. What are researchers most concerned with testing when they perform a study in a controlled laboratory setting?

Were Milgram's Obedience Studies Ethical?

For millennia, the field of philosophy known as ethics has studied the general nature of morals and morality. Morality refers to principles concerning the distinction between right and wrong (or good and bad) behavior. Thus, when concerned with the ethics of a research study, we are evaluating whether it would be right or wrong to perform the study. The evaluation is based on the harms ("costs") and benefits experienced by all involved: the subjects, the researchers, those who read about the study, etc. (See the Ethical Principles of Psychologists and Code of Conduct published by the American Psychological Association.)

In evaluating whether Milgram's research was ethical or not, one important question to ask is: would you have wanted to be a participant in one of his experiments? If you answered, "no," you probably were concerned about the distress felt by most participants as they struggled with the conflict between defying an authority and harming a human being. You also might have been disturbed by the possibility that some participants felt as if they had been made to look like “fools.” Neither of these harms would have occurred if deception (lying to participants about the purpose of a study) hadn’t been used. Elms (1995) described the four pieces of information withheld from the Teachers:

Number one, the experiment was a study of obedience to authority, not a study of memory and learning. Number two, the volunteer who assumed the role of learner was actually an experimental confederate. Number three, the only shock that anyone ever got was the 45 volt sample shock given to each teacher; the shock generator was not wired to give any shocks to the learner. Number four, the learner's kicks against the wall, his screams, his refusals to continue, were all carefully scripted and rehearsed, as were the experimenter's orders to the teacher.

But, if the participants had known this information, there could have been no study of obedience: deception was necessary if Milgram was to test his hypotheses about the situational factors involved in obeying an authority in an extreme situation (see Elms, 1982; Elms, 1998d). The dilemma we face pits moral concerns about causing others to experience distress, embarrassment, and other psychological harms against scientific concerns about how best to test causal hypotheses.

The use of deception may have been an important psychological harm for participants, and some have questioned the ethics of the study. Critics of Milgram’s studies have asked questions such as the following:

• Is scientific understanding of humans such an important goal that researchers have the right to deceive people, thereby causing psychological harm?
• How can we determine if researchers have other, more selfish, goals motivating them to perform their research?

With respect to the second question, perhaps some researchers simply want to satisfy their career ambitions. In this case, it is conceivable that they would choose research topics and design studies that would draw the most attention. The more extreme and controversial the research, the more attention the researchers will receive. In fact, Milgram’s obedience studies have become well known, not only by other behavioral scientists, but also by the general public. For instance, a play and a television movie based on his studies were produced; and a discussion of the obedience studies is a staple of introductory psychology courses around the world.

Perhaps the most important ethical problem for deception is this: when people are deceived, they may not be given all the information relevant for deciding whether or not to participate. The process known as informed consent involves telling potential participants about all aspects of the study relevant to making the decision to participate, including any physical or psychological harm that may result from their participation. One psychologist, Diane Baumrind (1964) argued that, rather than giving what we now call their informed consent, the participants had been "entrapped" into performing actions that caused psychological harm (see Elms, 1995). Milgram’s participants, when they first agreed to participate, did not know that the Experimenter would command them to seriously harm another person. If the Experimenter had mentioned this possibility, it is likely that most participants would have refused to take part in the study. During the 1970’s, an ethical requirement was established that stated that subjects must give their to participate in a study. This means that that potential participants must be told before agreeing to participate about everything that might be harmful to them (physically and psychologically). The ethical problem with deception should now be obvious: are people really able to give their informed consent when they are lied to about fundamental aspects of a study?

Milgram (1964, 1974) defended himself against such criticisms by pointing to the fact that almost all subjects had a favorable view of the study after they had participated in it. Milgram argued that this was the most important factor justifying the continuation of any study: “The central moral justification for allowing a procedure of the sort used in my experiment is that it is judged acceptable by those who have taken part in it.” In other words, he argued that "whether it is unethical to pursue truths [by creating an illusion, thereby deceiving people] ... cannot be answered in the abstract. It depends entirely on the response of those who have been exposed to such a procedure" (Milgram, 1974, p. 199).

Milgram (1974) stated that, if many of the participants had expressed their outrage at having been subjected to the experimental manipulations, he would have stopped the research. Milgram claimed that, rather than being outraged, many participants reported that the experience had been personally meaningful: “They viewed the experience as an opportunity to learn something of importance about themselves, and more generally, about the conditions of human action” (p. 196). His point was that deception and emotional upset are not the moral problems some critics have claimed them to be — not when the participants themselves believe their experiences to have been positive and meaningful. Furthermore, he took care to reassure participants as soon as the experiment ended. According to Elms (1995): "The experimenter gave each subject a standard debriefing at the end of the hour, to minimize any continuing stress and to show that the 'victim' had not been injured by the 'shocks'. When a subject appeared especially stressed, Milgram often moved out from behind the curtains to do an especially thorough job of reassurance and stress reduction."

A possible problem with Milgram’s line of reasoning involves an extension of a point made in Section 5-1: when people commit themselves to a course of action, they become motivated to think of their decision as having been the correct one. Extending this idea to the present case, once people have agreed to participate in a study, they become motivated to think of their participation as having been correct and justifiable, even if they experienced harmful consequences. In fact, there is research demonstrating that the more distress a person feels after agreeing to a course of action, the greater his or her motivation to think of the decision as having been correct (Aronson & Mills, 1959; Cialdini, 1998). It is thought that this helps people to maintain their self-esteem after making what might seem to be a poor decision.

Is the motivation to maintain their self-esteem the reason that almost all of Milgram’s participants evaluated the study as having been a positive experience? That is, was it just “wishful thinking” on their part or did they actually experience something deeply meaningful and positive? It is impossible to say. It even is possible that their motivation to see their experiences in a positive light caused them to focus on the positive aspects of the study and to ignore its negative aspects, which would make their experiences positive and meaningful even if this was not their original view. Thus, it may depend upon how others describe the experience, especially authorities who help the participants to interpret their experiences. If this is true, then Milgram's “debriefing” of participants immediately after the experimental session ended, which now is an essential component of any study of humans, but was novel in 1961, helps participants to interpret their participation positively, thereby reducing potential psychological harms.

In addition to the meaningfulness of the experience for the participants, there are other benefits of research. The most important benefit is the knowledge gained from analyses and interpretations of a study's results. In fact, because of the importance of Milgram's findings to our knowledge of human nature, here you are, decades after his research was completed, reading about it in a textbook. In general, when we are evaluating the ethics of a proposed study, we must weigh its potential benefits against it potential harms (costs). If the benefits outweigh the costs to a sufficient degree, the study is judged to be ethical. But there are difficulties inherent in making this judgement. First, to what degree must the benefits outweigh the costs before we judge the study to be ethical? There is no precise and unambiguous answer to this question. Second, which standards should be used to assess the amount of benefit or harm produced by a research study. It seems obvious that each person making such an assessment will rely on his or her own standards (values). For example, some people may feel that it is always wrong to cause another to experience distress or that scientific knowledge is not important. Others may feel that scientific knowledge is of paramount importance or that emotional suffering is good for one's character.

Regardless of your own judgement about the ethics of Milgram's research, it no longer is possible to replicate his studies in the United States. Most colleges and universities in the United States have “institutional review boards” (IRBs) that consider the ethics of proposed studies. It is very unlikely that an IRB would approve any study similar to Milgram’s. Some applaud this fact whereas others feel its stifling effects upon the science of psychology (see Illinois White Paper, 2005). In fact, some researchers believe that it is impossible any longer to do any truly important research in social psychology.

Study Questions

15. Why is deception used in many social psychological studies?
16. What is(are) the ethical problem(s) with respect to the use of deception? (In your answer, please discuss the issue of informed consent.)
17. Why did Milgram not think that his use of deception and the extreme distress experienced by many of the participants violated ethical standards?
18. Why is it important to have debriefing sessions for participants after they have completed an experiment?
19. What are possible benefits of research studies?
20. How do we decide whether or not a study meets ethical standards for research?
21. Why do people disagree about whether a study is ethical or not?
22. What are institutional review boards (IRBs)?
23. Why do many research psychologists criticize IRBs?

How Are We Influenced By Social Roles?

Prisons are places in which all kinds of brutal behaviors are a matter of everyday routine. Is this brutality the result of a large number of psychopathic criminals and sadistic guards all concentrated in one place? That is, is such behavior caused by the unique personality characteristics exhibited by the people who are found in prisons? After reading about Milgram’s obedience study, you probably are beginning to realize that there may be other explanations for such brutality. You probably now could suggest the possibility that the social situations found in prisons are defined by the participants in certain ways — ways that cause them to act in a more brutal manner. Thus, just as in Milgram’s study, people in prisons may adopt particular social roles that cause them to act, think, and feel in particular ways.

In 1971, Philip Zimbardo and several colleagues tested this claim (see The Stanford Prison Experiment). Their main goal was to see the degree to which the social roles adopted in prisons affect the behaviors and mental events of people. The idea for the study originated in an undergraduate social-psychology course taught by Zimbardo at Stanford University. He divided the class into study groups of about ten to fifteen students and asked them to investigate various topics. One of these topics involved the effects of prisons on the behaviors of those who lived or worked within them. The group given this topic decided to investigate it by simulating a prison environment. Thus, over the course of one weekend, they maintained a mock prison in the basement of a student dormitory. Some of the students played the part of prisoners and others became guards. The simulation worked very well, perhaps too well. The people who were playing the role of guards seemed to take their roles very seriously. In fact, they behaved very aggressively towards the “prisoners.” Zimbardo (1975) stated that, by “the end of the weekend some long-term friendships were broken because those young men and women who were prisoners believed that in their roles as ‘mock’ guards the ‘true’ self of their former friends were revealed, and they could no longer befriend such sadistic authoritarian people” (p. 37).

Zimbardo and the students could not understand what had happened. Why had this simulation, this “game,” become so serious? Was there something unique about the individuals who made up this group — some personality characteristic that caused them to act sadistically—or was there something about the social situation being simulated that would cause most people to act in similar ways? Zimbardo and several of his colleagues tried to develop an internally valid study that might give them answers to these questions. Their subjects were college students from around the country, many of whom were visiting the San Francisco area for the summer. Several measurements of their personalities were made with respect to relevant attitudes, past behavior, and past experiences. Based on these responses, Zimbardo and his colleagues chose only those subjects who showed no obvious tendencies toward criminal or violent behavior, and no evidence of serious mental illness. They then set up a simulated prison in the basement of the psychology building at Stanford University. The subjects were paid $15 per day for their participation in a simulation that was to last two weeks. Zimbardo and his colleagues argued that, because they had excluded those subjects who had abnormal personalities and had randomly assigned participants to be either guards or prisoners, they would be able to conclude that any violent or sadistic behavior that occurred in the simulated prison was caused by the social context itself.

When first learning about this study, students sometimes argue that Zimbardo and his colleagues should have studied actual prisoners and guards in an actual prison situation. These students state that the researchers could simply have asked the prisoners and guards why they acted the way they did in particular situations. But think carefully about this. Why would such a study, which definitely would have high external validity (because it would take place in a natural situation), not be an internally valid one? It would not be internally valid because people who already are prisoners and guards might exhibit the personality characteristics that we would want to exclude in any study investigating the ability of the social situation to affect behaviors and mental events. Furthermore, asking actual prisoners and guards why they acted as they did would not be an internally valid procedure because they are not likely to be aware of the situational influences on their behavior.

Zimbardo and his colleagues attempted to simulate the natural prison environment as closely as possible because they did not know which factors in that situation might be important for the behaviors they wanted to study. Thus, all subjects who were to play the role of prisoner were told to be home at a certain time, and then they were arrested by officers from the Palo Alto City Police Department. Many of the subjects were not certain at first whether this was an actual arrest or a part of the study:

A police officer charged them with suspicion of burglary or armed robbery, advised them of their legal rights, handcuffed them, thoroughly searched them (often as curious neighbors looked on) and carried them off to the police station in the rear of the police car. At the station they went through the standard routines of being fingerprinted, having an identification file prepared and then being placed in a detention cell. [In the only difference from an actual arrest, each] prisoner was blindfolded and subsequently driven by one of the experimenters and a subject-guard to our mock prison.... Upon arrival at our experimental prison, each prisoner was stripped, sprayed with a delousing preparation (a deodorant spray) and made to stand alone naked for a while in the cell yard. (Haney, Banks, & Zimbardo, 1973, p. 4).

In order to make all subjects feel less like unique individuals, they were dressed in uniforms befitting each role: “For the guards, the uniform consisted of plain khaki shirts and trousers, a whistle, a police night-stick..., and reflecting sunglasses.... The prisoners’ uniform consisted of a loose fitting muslin smock with an identification number on front and back, no underclothes, a light chain and lock around one ankle, rubber sandals and a cap made from a nylon stocking” (p. 4). The uniform of the prisoners was designed specifically to eliminate their masculine identity: “The prisoner uniforms resembled smocks or dresses, which made them look silly and enabled the guards to refer to them as ‘sissies’ or ‘girls’. Wearing these uniforms without any underclothes forced the prisoners to move and sit in unfamiliar, feminine postures” (p. 8).

The ten prisoners who participated in the study were kept in the prison 24 hours per day, whereas the eleven guards had eight-hour work shifts (three guards for each shift) and then could leave to go back to their own homes. The prisoners were housed in three “cells” (three per cell) and had no privacy during the entire length of the study. In his instructions to the guards before the study began, Zimbardo stressed the fact that physical abuse of the prisoners would not be tolerated. Other than this, however, he gave few instructions to them about how they were to behave: he wanted to see how deeply they would get into the roles on their own. The prisoners had been told before the study began that, although some of their basic civil rights would be suspended during the study, they would not be physically harmed and they would have an adequate diet, clothing, medical care, and housing. Again, other than this, he gave them little information.

Once in the prison, the prisoners quickly became very distressed by their treatment. This distress was seen in their negative emotions, a generally negative outlook, and negative evaluations of themselves and their environments. In fact, five prisoners had to be released within several days after the study began, four of them for severe depressive symptoms such as crying, rage, and anxiety (the remaining prisoner developed a severe rash that may have been due to stress). Guards also showed more negative emotions, self-evaluations, and general outlook. Nevertheless, most of them also seemed to enjoy the power of their roles. In fact, they were always on time for their shifts, and would sometimes stay for several hours after their shifts were over without extra pay and without complaint. Some individual differences that seemed to reflect the unique personalities of the participants emerged among both the prisoners and the guards.

There seemed to be three kinds of guard: some acted sadistically by trying to humiliate the prisoners; some enforced the rules in a tough manner but did not often try to humiliate the prisoners; and some exerted almost no control over the prisoners and would sometimes even do them small favors (even these guards, however, sometimes acted like the more sadistic guards). There also seemed to be more than one kind of prisoner, although these differences weren’t as obvious since almost all prisoners, after the first few days, became passive, depressed, and obedient. Nevertheless, two prisoners were consistently rebellious (one of whom was discharged on the second day because of an extreme emotional reaction to the simulation), about half developed symptoms of emotional disturbance, and about half became excessively obedient. Although these patterns of responding to the prison situation, especially among the guards, seemed to reflect personality differences, these differences all occurred within the confines of the social roles assigned to the subjects. That is, not one subject ever stepped outside of the social role he had voluntarily accepted. For example, although there were guards who seemed to feel sorry at times for the prisoners, and would even try to help them occasionally, none of them ever intervened to stop the humiliating tactics of the more sadistic guards. In other words, all participants accepted the particular social roles they were assigned, and defined themselves and everyone else according to these roles.

Perhaps the most disturbing reaction was an increase in group cohesiveness among the guards (they tended to stick together with respect to their abuse of the prisoners) and a disintegration of group cohesiveness among the prisoners. Because of this, the guards could act as one unit of oppression whereas each isolated prisoner, lacking any hope of effective rebellion (because he would not be supported by the other prisoners) had almost no choice but to submit to the overwhelming authority of the guards. In fact, many of the prisoners even sided with the guards when a particular prisoner rebelled by refusing to eat. As a consequence of their virtually absolute power, the guards became increasingly abusive, and the prisoners became increasingly passive. One detail of the increasingly aggressive behaviors of the guards mirrors something we saw in Milgram’s study:

many guards continued to intensify their harassment and aggressive behavior even after the second day of the study, when prisoner deterioration became marked and visible and emotional breakdowns began to occur.... When questioned after the study about their persistent affrontive and harassing behavior in the face of prisoner emotional trauma, most guards replied that they were “just playing the role” of a tough guard, although none ever doubted the magnitude or validity of the prisoners’ emotional response. The reader may wish to consider to what extremes an individual may go, how great must be the consequences of his behavior for others, before he can no longer rightfully attribute his actions to “playing a role” and thereby abdicate responsibility. (Haney, Banks, & Zimbardo, 1973, p. 6)

What is surprising about the prison study, and what makes it different from Milgram’s study, is that the guards abused the prisoners even though no authority figure was present to encourage or command them to do it. Once they had defined themselves as being the relevant authorities in the situation by adopting the role of “prison guard,” they felt constrained by these roles to act accordingly. In fact, the guards saw the researchers in charge of the study as being too soft, and tried to conceal some of their own more abusive behaviors from them. The moral of the Stanford Prison Study seems to be this: when a social situation becomes “internalized,” when it is “brought inside” of a person in such a way that it affects how that person looks at his/her self and his/her surroundings (including other people), the person’s role in that situation determines much of his/her behavior, thoughts, and emotions.

This moral of the study is illustrated by the diary entries of one person who had been assigned the role of a guard. Before the study, he had written that he could not “see a time when I might ... maltreat other living things” because he thought of himself as opposed to the use of coercive force under any circumstances. He also thought that the study was just going to be a fun game that no one would take seriously. After the study began, however, his thinking and behavior quickly changed:

First day:  Feel sure that the prisoners will make fun of my appearance and evolve [decide upon] my first basic strategy — mainly not to smile at anything they say or do which would be admitting it’s all only a game....

Second day:  Number 5704 asked for a cigarette and I ignored him — because I am a nonsmoker and could not empathize.... Meanwhile since I was feeling empathetic toward #1037, I determined not to talk with him.... After we had count and lights out [another guard] ... and I held a loud conversation about going home to our girlfriends and what we were going to do to them.

Third day:  (Preparing for the first visitors’ night.) After warning the prisoners not to make any complaints unless they wanted the visit terminated fast, we finally brought in the first parents. I made sure I was one of the guards on the yard because this was my first chance for the type of manipulative power that I really like — being a very noticed figure with almost complete control over what is said or not....

Fifth day:  I harass “Sarge” [an extremely obedient prisoner] who continues to stubbornly overrespond to all commands. I have singled him out for special abuse both because he begs for it and because I simply don’t like him. The real trouble starts at dinner. The new prisoner (416) refuses to eat his sausage ... we throw him into the Hole ordering him to hold sausages in each hand.... We decide to...tell the new one that all the others will be deprived of visitors if he does not eat his dinner.... I am very angry at this prisoner for causing discomfort and trouble for the others. I decided to force-feed him, but he wouldn’t eat. I let the food slide down his face. I didn’t believe it was me doing it. I hated myself for making him eat but I hated him more for not eating. (Zimbardo, 1975, pp. 48-49)

These entries suggest that this person no longer felt any empathy for the prisoners. He now seemed to identify completely with his role as a prison guard. Zimbardo concluded that everyone, including himself as the “prison superintendent,” had gotten lost in their roles—their identities had become their social roles. This process led to increasingly abusive behavior on the part of the guards and almost complete passivity, as well as severe emotional disturbances, on the part of the prisoners.

Zimbardo observed all this abuse and initially thought that nothing was wrong. In fact, he saw all of it as proof that the study was a success. He reported that he did not become aware of how bad the situation was until his fiancée expressed how upset she was about the sadistic treatment the prisoners were experiencing. At that point, Zimbardo reluctantly called a halt to the study after only six days of what was to be a two-week study:

At the end of only six days we had to close down our mock prison because what we saw was frightening. It was no longer apparent to most of the subjects (or to us) where reality ended and their roles began. The majority had indeed become prisoners or guards, no longer able to clearly differentiate between role playing and self. There were dramatic changes in virtually every aspect of their behavior, thinking and feeling. In less than a week the experience of imprisonment undid (temporarily) a lifetime of learning; human values were suspended, self-concepts were challenged and the ugliest, most base, pathological side of human nature surfaced. We were horrified because we saw some boys (guards) treat others as if they were despicable animals, taking pleasure in cruelty, while other boys (prisoners) became servile, dehumanized robots who thought only of escape, of their own individual survival and of their mounting hatred for the guards. (Zimbardo, 1972; reprinted in McIntyre, 1999, p. 114)

The guard whose diary entries were quoted above stated that he was “elated” that the study was ended early, but that he also was “shocked to find some other guards disappointed somewhat because ... they are enjoying themselves” (p. 49). This person did not seem to remember that he also had been enjoying his power, and using this power in a manner meant to humiliate the prisoners. Somehow, he was able to act in sadistic ways and still see himself as a pacifist.

Study Questions

24. What inspired Philip Zimbardo to perform the Stanford Prison Study?
25. What hypothesis was Zimbardo and his colleagues testing?
26. Why did they not use an actual prison situation to test this hypothesis?
27. How did they control for the effects of personality?
28. What did they do to make their prison simulation as real as possible?
29. What did they do to make sure that the participants became immersed in their assigned roles?
30. What individual differences emerged in those assigned to the prisoner role?
31. What individual differences emerged in those assigned to the guard role?
32. What happened with respect to "group cohesiveness" in the prisoners and in the guards?
33. In what way did the cruel behaviors of the guards in Zimardo's study differ from the cruel behaviors of the Teachers in Milgram's study?
34. Why did this difference occur between the two studies?
35. In what ways did the thinking of the prison guard who kept a diary change over the course of the experiment?
36. Why did Zimbardo end the study over a week before it was supposed to end?

Comparing Milgram's Obedience Study to Zimbardo's Prison Study

There are several similarities between Zimbardo’s prison study and Milgram’s obedience study. First, subjects in both studies had a difficult time ending their participation, and most continued all the way until the end. The reasons for this were the same in the Zimbardo study as in the Milgram study.

1. By agreeing to take on the roles they had been assigned, it became very difficult for the subjects to back out without breaking the implicit social rule I spoke of earlier.
2. In addition, they did not want to appear inconsistent by refusing to continue something they had already started.
3. Lastly, the escalation in violence occurred relatively gradually, so that there was no obvious point at which they could say, “that is enough.”

A second similarity to the Milgram study was the limited effect of measurable personality differences on the behavioral differences of the prisoners and guards. Again, the only personality characteristic that seemed to have any effect was authoritarianism; and this characteristic was important only for prisoner behavior. It seemed that those prisoners who were high in authoritarianism were best able to handle the oppressive conditions in the prison and, thus, remained there the longest. At least four (and maybe five) of the initial group of nine prisoners, on the other hand, had to be released even before the study was ended because of severe stress caused by the conditions.

Based on results such as these, Zimbardo argued that personality differences were much less important than the social situation. In fact, Zimbardo went even farther than Milgram had by dismissing the importance of personality characteristics for most of our behaviors in everyday life:

Individual behavior is largely under the control of social forces and environmental contingencies rather than personality traits, character, will power or other empirically unvalidated constructs. Thus we create an illusion of freedom by attributing more internal control to ourselves, to the individual, than actually exists. We thus underestimate the power and pervasiveness of situational controls over behavior because: (a) they are often non-obvious and subtle, (b) we can often avoid entering situations where we might be so controlled, (c) we label as “weak” or “deviant” people in those situations who do behave differently from how we believe we would. (p. 115)

Many social psychologists took this extreme position during the 1970’s. In the next chapter, I will discuss why they argued for this position and will try to show why it needs to be modified and softened somewhat. In fact, I would imagine that Zimbardo himself would no longer make such a strong claim, although he and many other social psychologists still give social forces a large share of the control over our behavior and mental events.

A major difference between the prison study and Milgram’s obedience study was that, in the prison study, only one experimental manipulation was performed — being assigned to the role of prisoner or guard. This is a major limitation because we cannot know, as we did in the Milgram study, which factors in the situation were most important for the behaviors observed. There is no doubt that each subject’s self-definition as prisoner or guard was important, but it might have helped us to develop a deeper understanding of how such a self-definition can be maintained if Zimbardo and his colleagues had varied other factors in the situation. For example, would wearing normal clothes, which might have caused increased feelings of individuality, have resulted in decreased role-playing? Would a decrease in the reality of the simulation (perhaps by removing the bars on the doors, or having campus security “arrest” the prisoners) have done the same? It would have been very interesting, for example, if we had found that none of these factors were important — that, instead, it was simply the assigning of an arbitrary social role by an authority figure and the voluntary taking on of that role by the subject that was most important for his role-playing behavior. We probably never will be able to find out because, as I will discuss in the next section, it is unlikely that anything similar to this study could ever be performed today.

Zimbardo stopped the study because he felt that it was unethical to continue. In fact, years later, he was to state that the study was unethical as a whole. Based on the discussion of research ethics above, what do you think made the study unethical? Probably the most important aspect of the study making it unethical was the psychological harm that was done to the prisoners. For example, the prisoners were depressed and pathologically obedient because of the punishments (mostly loss of everyday privileges such as being able to go to the bathroom, eat, sleep, and so on) that were imposed by the guards. Although the argument has been made that Milgram’s study also caused much psychological harm, the distress felt by the teachers in that study was temporary (the procedure lasted only an hour), and Milgram showed that almost all had either positive or neutral feelings about the study afterwards. The prison study, on the other hand, led to very negative feelings among most of the subjects that lasted for at least several months. They were brought back on several occasions after the study ended to discuss their feelings and thoughts about the study. It is evident from their statements that many or all of them were bothered very much by their own behavior and by the behavior of the other subjects. Another aspect that, in the present day, is considered to be unethical was the dual role of Zimbardo, who was both the “prison superintendent” and the principle researcher of the study. Taking on such a dual role made it difficult for him to monitor the progress of the study objectively. He stopped the study only reluctantly when it was pointed out to him that the prisoners were suffering terribly. In the obedience study described earlier, Milgram was never present during the actual procedure itself, and so he was able to be somewhat more objective about any harm being done to the subjects.

Study Questions

37. In what ways were the results of Zimbardo's and Milgram's studies similar?
38. Why did they get such similar results?
39. What was Zimbardo's main conclusion?
40. What makes this conclusion so extreme?
41. What was a major limitation of Zimbardo's study with respect to discovering the factors that were most important for the influence of social roles on behavior?
42. Why did Zimbardo not realize initially that his study was unethical — that its harms outweighed its benefits?