Simulated Prison


Most of us know about Abu Ghraib and the appalling acts carried out by American soldiers. If not, do a bit of research but, be warned, the information is extremely graphic and disturbing.

So, when you think about these soldiers torturing prisoners, you might think, “What an evil person…” and I would nod my head in agreement. But then again, why did that soldier behave so inhumanely? Were they born evil? A lot of researchers believe so, bringing up the ‘dispositional hypothesis‘ (i.e. the guards behaved in an evil manner because they have evil personalities). Professor Zimbardo carried out this controversial experiment – funded by the US Navy and Marine Corps – in the early 70s, hoping to test this hypothesis. When he was a young boy, he saw his friends engaging in violence, drugs and crime. He was fixated on the concept of good people doing bad things, and believed that the environment was the biggest influence. Zimbardo set out to prove that guards are not naturally sadistic and that prisoners are not naturally aggressive, but that their situation (i.e. a deplorable, disturbing prison) provokes their behaviour: the ‘situational hypothesis‘.

He also discussed the social roles given to guards and prisoners. Social roles can be expected from everyone, such as students like you who are expected to work hard and revise. (Oh wait, you are revising! Good job!)

Is it a bird? Is it a plane? No! …It’s a psychologist.

A newspaper advertisement for male volunteers willing to take part in a study on “prison life” was sent out, offering $15 per day of the experiment. 75 men responded and completed a questionnaire about their family background, mental health, physical health, past experiences and their attitude in relation to crime and psychopathology. After this, a final 24 was selected. This sample was concluded to be the most physically able, mentally stable and mature. They were also the least antisocial. They were described as “normal, healthy male college students”. The majority of them were White and middle-class.

Credit: Philip G. Zimbardo, Inc.

The sample was to be separated into ‘guards’ and ‘prisoners’. All participants signed a contract which guaranteed them basic living needs but the prisoners were told that some of their civil rights would be suspended during the experiment. No further information was given, expect instructions telling them to be available at their homes on the Sunday of the experiment.

The men who had been given the role of guards attended an orientation meeting the day before prisoners were involved. The guards met the principal investigators, the superintendent (Zimbardo!) and the warden (an undergraduate research assistant). Their task was to “maintain the reasonable degree of order within the prison necessary for effective functioning” without aggressive physical punishment. They were also given further details on their work-shifts, the writing of critical incident reports, meal management, work and recreational prisoner programmes. The guards then assisted in building the make-shift prison (a room at Stanford University’s psychology department). The guards were under the impression that the experimenters were interested in studying the behaviour of prisoners.

All the guards and prisoners were given a distinctive uniform. Guards wore a plain khaki shirt and trousers, reflective sunglasses (think Terminator), a whistle and a wooden baton. Prisoners wore a loose smock with an ID number on the front and back, no underwear, rubber sandals, a hat made from nylon stockings and a light locked chain around their ankle. Can you imagine being a normal college man and having to wear one of these uniforms? The guards’ uniform was designed to give them a military-like persona of power, control and intimidation. The prisoners were made to feel emasculated (less masculine), uncomfortable and depersonalised (lacking a personal identity). Their ID number was used instead of their names.

One Sunday, the prisoners were unexpectedly arrested at their homes. A police officer charged them with suspicion of buglary or armed robbery, read them their rights (you know, they have the right to remain silent and so on), handcuffed them, searched them and drove them away in a police car. This was in full view of neighbours and people passing by, making the situation very realistic and probably embarrassing.

They were taken to a police station, gave their fingerprints and took mugshots. Each prisoner was blindfolded and driven to the make-shift prison. During all this, the police officers were serious and formal. At the make-shift prison, the prisoners were stripped, sprayed to get rid of germs (gee, if the stripping wasn’t embarrassing enough) and left to stand alone and naked (well… that escalated quickly). After putting on their uniform, they were put in a cell and told to remain silent.

The warden read the institution rules which the prisoners had to memorise and follow.

The prisoners were given three plain meals, three supervised toilet visits and two hours of reading or letter writing. Three times a day, they were lined up for ten-minute counts which included testing them on the rules and their ID numbers. As the prison situation became worse, the ten-minutes increased to a point where counts were lasting for hours instead of minutes.

The guards would rotate their shifts so when they were outside the prison they would resume their normal lives as students who study and socialise. The prisoners, on the other hand, remained in the prison 24 hours a day. The experiment was supposed to last for two weeks, but only lasted for 6 days.

The experiment contradicted the dispositional hypothesis. How? I’m about to tell you.

As stated previously, the experiment was terminated after only 6 days because the state of the prison deteriorated so fast. Five prisoners were released early because of extreme depression. They had been crying, anxious and angry. The prisoners were relieved and visibly happy at the end of the experiment but the guards seemed to be distressed and disappointed that it had ended.

The participants were all becoming very negative. Even interactions between guards and prisoners were hostile and awkward. They had begun to really believe in their roles and adopted very distinct behaviours. Guards were commanding while prisoners were passive. Individual differences were also seen: half the prisoners endured the oppression while others were a little more rebellious, and some guards were tough but fair while others went beyond their role to engage in cruel harassment, including a ban on toilet privileges and forcing prisoners to urinate in a bucket instead (a bucket which stayed unclean and remained in the cell). Some guards even worked extra hours without pay.

The prison was a miniature reality with its own laws, rebellions, rumours and conflict. Remember, the participants were free to withdraw from the study at any time but none of them left as a choice. They either became so disturbed that they were let out by the experimenters, or stayed until the end. Also, after the first five hysterical prisoners were let out, only two of the remaining prisoners said they were not willing to lose the money in exchange for “parole”. The ones who applied for parole (meaning they lost the money) were rejected, and yet they STILL didn’t walk out. Even when a priest visited the prison, the prisoners referred to themselves with their ID number instead of their real name and some asked the priest for a lawyer to help them get out of the prison.

Even Zimbardo, the head psychologist, became wrapped up in the experiment. He visited the make-shift prison and let it continue even when the abuse and humiliation were in front of his eyes. The reason the experiment ended was actually because Zimbardo’s girlfriend Christina Maslach (now his wife, aww!), a graduate student at the time, saw the state of the prison and objected. It should be noted here that out of the 50+ people who saw the conditions of the make-shift prison, Christina was the first and only person to complain.

Credit: Philip G. Zimbardo, Inc

The guards enjoyed their power and abused it from near the beginning of the experiment, and Zimbardo attributed this behaviour to a “pathology of power”. The prisoners became emasculated, helpless, dependent and depersonalised. Zimbardo called this the “pathological prisoner syndrome”.

Type of research method
Zimbardo’s study can be seen as a field experiment. It was not conducted in a laboratory but we can look at it as a field experiment because it was a relatively non-artificial environment and certain factors were manipulated (such as the roles).

Independent variable
The role (i.e. prisoner or guard) allocated to a participant is the IV, as well as the environmental conditions (i.e. prison and cells).

Dependent variable
The behaviour of the participants is the DV.


  1. Some ecological validity: The experimenters created a make-shift prison, put prisoners through a true-to-life arrest procedure and conducted activities that are common in real-life prisons. This increases the ecological validity and makes the situation more realistic.
  2. Data collection: Zimbardo used a variety of data collection methods, such as: interviews, observation (covert and overt) and questionnaires. This means a large amount of data can be compiled together, and most of it would be in-depth qualitative data.
  3. Quite replicable: Although it was a very complex experiment, some parts of it can be replicated because they were standardised. This includes the questionnaires, the arrest procedure and the prison environment.


  1. Not generalisable: The sample size was small (only 24 people) and consisted of predominantly White, middle-class, American men. Therefore, the results cannot be applied to women, non-Americans, a different socio-economic background or other ethnicities.
  2. Selection bias: The participants all responded to an advertisement about taking part in an experiment about “prison life”, meaning they were probably comfortable with an idea that other people may be uncomfortable with. Although the participants were volunteers, the selection process was not totally random.
  3. Demand characteristics: Since Zimbardo was part of the experiment as a superintendent, he may have influenced the participants unintentionally and affected their behaviour or attitude.

Ethical issues

  1. Informed consent: Zimbardo did get informed consent from the participants.
  2. Deception: There was not any deception in this study, although some of the facts were given in such a way that the participants made their own conclusions about the experiment (like the guards thinking the experiment was more about prisoner behaviour).
  3. Confidentiality: The participants’ identities were not made public for the world to see, but eventually a lot of them talked about the experience so some of their identities were known, although not because of Zimbardo.
  4. Emotional or physical harm: While there were no physical injuries, there was a whole bucketful of emotional harm. Many of the prisoners were severely disturbed and upset by the study, especially the ones that left early. However, Zimbardo did take measures to make sure they felt better afterwards.
  5. The right to withdraw: All participants were allowed to withdraw at any time, although none of them withdrew.
  6. Debriefing: Group sessions and individual meetings for the sake of debriefing took place after the experiment. Participants also took questionnaires afterwards and kept in further contact with yearly intervals.

Reference: Haney, C., Banks, C. and Zimbardo, P. (1973). A Study of Prisoners and Guards in a Simulated Prison. Naval Research Reviews.