The paper “A Study of Prisoners and Guards in a Simulated Prison,” originally published in 1973 and reprinted in 2002, describes a study popularly known as the Stanford Prison Experiment. It studies the rehabilitation capabilities of American prisons. Due to the recidivism rate of 75%, there’s the belief that US prisons do not rehabilitate criminals. The fact that, by 1973, the Americans were investing 1.5 billion dollars in the correction system makes the problem a significant one to study. The authors had an assumption about the environment of the prison: that it causes people to commit despicable actions, and, by consequence, changes their behavior significantly. Strictly speaking, the hypothesis of the study was that, after a period of two weeks under a prison environment, people from the same group would present “significantly different reactions on behavioral measures of interaction, emotional measures of mood state and pathology, attitudes toward self, as well as other indices of coping and adaptation to this novel situation” (Haney, Banks, & Zimbardo, 2002, p. 4).

To sample the population, the researchers published a newspaper ad inviting adult males for interviews. From a pool of 75 people who showed up, they removed all the ones with history and propensity for psychopathologies. Later, they selected a group of 24 men, the ones whom they considered most physically and mentally stable from the original sample. As payment for participation, they offered each of them a daily amount of $15. The researchers would build a mock prison under the boundaries of the Psychology building at Stanford University. Although it didn’t have the intention to create an accurate simulation of a prison, they expected to represent one. The people would stay 24/7 in this fictional penitentiary, randomly designated to act as either as a guard or a prisoner.

From the 24 selected men, the experiment started with 11 guards and 10 prisoners. Two other men stayed at home, in a standby role in case of need, and one person gave up on participating right before starting it. The scientists told the “guards” they wanted to simulate “a prison environment within the limits imposed by pragmatic and ethical considerations,” (Haney et al., 2002, p. 7) without specific instructions on how to maintain order. “Prisoners” did not get instructions on how to behave in the controlled environment either.

For later experimental analysis, the data collected to measure dependent variables were of two types: “transactions between and within each group of subjects, recorded on video and audio tape as well as directly observed”; and “individual reactions on questionnaires, mood inventories, personality tests, daily guard shift reports, and post experimental interviews.” (Haney et al., 2002, p. 5). As previously mentioned, the only independent variable expected by the researchers was the role assigned to the subject, either guard or prisoner.

According to the authors, the results were consistent with reports made by ex-convicts. Both groups under test, on average, showed increased levels of negativity and aggressive behavior. Prisoners expressed signals of extreme emotional depression, and guards seemed to enjoy the power they had over others. The experiment started to become so dangerous for the health of the subjects that they decided to abort it after only six days.

In this short period, the prisoners showed signals of loss of personal identity, significant changes in their behavior due to arbitrary control from the guards, and the constant need to pose as aggressive ones over other prisoners. This result, assuming it could be replicated, shows the dangerous consequences of incarceration. When a penalty is intended for rehabilitating a human being, the criminal system should not cause psychological harm over prisoners or guards. For this reason, this experiment suggests that more studies should take place in this area of knowledge.

Although highly dangerous, according to the authors, the experiment was capable of putting people under circumstances very similar to being incarcerated for real. They presented significant differences between subjects before and after the situation, where guards were even working extra hours for free and, later, having prisoners expressing the intention of causing harm to their peers. This conclusion, according to the publication, presents the experiment as consistent with real environments they were set up to replicate, as attested by a prison consultant and a priest who were in direct contact with both subjects and real prisoners on other occasions.

On the ethical side, this experiment would probably have not being approved if submitted nowadays. An Institutional Review Board (IRB) must currently approve every experiment involving humans, and they are expected not to cause harm to the people under study. Considering this, the chief scientist did not stop the operation, even with people under psychological breakdowns. Also, there’s no mention of subjects being told about the possibility of leaving the mocked prison. Under the stress of the situation, people seemed to have forgotten about their rights and expectations for the experiment. These issues, impacting the subjects over multiple days, may have caused long-term consequences to the people who participated in the study.

On the scientific side, I can cite three substantial problems with the experiment. The first is that everyone under test knew they were part of a simulated environment. For this reason, it can never really reproduce a real prison, where subjects would not know how much time they would be deprived of freedom and under of high odds of experiencing violent attacks from others. Also, since everyone was aware of the experiment, there’s the chance that they were not behaving as they would normally do, but role-playing. One way of controlling this variable – role-playing or not – is not to tell the subject about the real intention of the study and employ actors for the other roles. Another problem with the experiment is that since it involved the interaction of multiple people, in the same environment, correctly controlling all the variables becomes quite difficult, especially with small sample size. The existence of a control group is not only useful to have but necessary to take any conclusions about the study.

Overall, my interpretation is that this study, as is, should not have happened. It has probably caused a negative impact over the mind and behavior of the people under research while additionally not returning reliable results for scientific purposes. I believe that modern ethical committees would agree with my interpretation.

This 1973 study tries to reveal that even a simulated and short experiment is capable of quickly dehumanizing the ones involved. Not only the prisoners could have had lasting consequences, but guards who saw themselves committing violent acts against other human beings. If there’s any validity, it shows that prisons from 46 years ago are not different from many we still see today, raising red flags over current public policies. It makes me interested in reading recent studies trying to reproduce the experiment in current prisons’ environments. Quite known in the Psychology field, the study brings awareness to the importance of electing politicians who rely on scientific studies for proposing law changes.


Haney, C., Banks, W. C., & Zimbardo, P. G. (2002). A Study of Prisoners and Guards in a Simulated Prison. Naval Research Review, 1–17.