Zombardi's prison experiment seems to provide very good evidence for how groups and societal influences can cause extreme behavior. The simulated prison in which the experiment was carried out had many characteristics of real prisons. Also, the assignment of students to either the role of guard or prisoner was totally random and test results showed that there were no initial difference between these two groups. These test results may have been unreliable. However, as the two groups were selected randomly, then it would be probable that the results were accurate. The cooperation of the Palo Alto Police Force to arrest the students who were going to be prisoners also helped to make the study realistic within the context of the experimental situation.
Even Zimbardo and his co-workers did not anticipate the extreme transformations in the behavior and attitudes that took place among the guards and prisoners in the simulated prison. They had expected group effects, but not the rapidity and intensity with which they developed. Zimbardo said that by the sixth day it had gone past the point of being an experiment and it had, in fact, become a prison for all the participants.
It can, however, be argued that the subjects who were placed in the positions of prisoners and guards were simply acting like good students and doing their best at role-playing and acting in the way they thought their parts required. If this were the case then this would be counter to the objectives of the experiment as it was not designed to discover whether the students were good actors. Also, the students who took the part of the prisoners were not actually criminals and may have experienced more stress than real criminals would have in the same situation. This may have been why they 'cracked' so early on.
However, there is a lot of evidence to suggest that it was group norms that affected the behavior of the subjects. Guards who did not seem to want to be brutal to the prisoners were pressurized into the use of abusive behavior by other guards. Also, prisoners who behaved independently were branded as troublemakers by the rest of the prisoners.
It's important to remember that experimental conditions are never exactly like real situations. However, in this case Zimbardo's prison simulation was extremely good. A big factor in this experiment which was different from a an actual prison situation is that the participants knew they were to be imprisoned for a relatively short time only. This does not seem to have mattered. The experiment was abandoned after six days as it had become almost exactly like a prison for the participants after just this short amount of time. Even the behavior of some of the experimenters themselves (who were serving as staff), including Zimbardo, was also affected. This helps to show the extent to which the participants internalized their roles (or felt them to be real). In these regards, the impact of the experimental situation being different from real prisons does not seem likely to have had much effect on the results.
Overall, although no experiment such as this can offer conclusive evidence about expected behavioral patterns the evidence for group behavioral patterns that this experiment provides is very convincing.
There were, of course, other influences on the behavior of the experiment participants. The prison-based nature of the experiment means that many other influences are likely to have affected the behavior of the subjects. Stereotypical norms associated with roles they were given were likely to affect the behavior of both groups. This conformity to stereotypes, along with power issues and consensus seeking behavior, will all have contributed to the way the experiment turned out. These influences seem to have largely had their expected effect, however. The guards and prisoners fell into stereotypical behavioral patterns and the imbalance of power between the two groups predictably led to the abuse of power on the side of the guards and the submission of the prisoner group. Consensus seeking behavior in both groups also seems to have contributed to the outcome. There wasn't much division within the groups and any dissenters were quickly drawn back into line. The deindividuation of the prisoners, just like in a real gaol, also seemed to help break them down quickly and push them into predictably submissive roles. Again, this would be a predictable outcome from that process.
It must not be forgotten that there are ethical criticisms which are often made against experiments such as Zimbardo's. In fact, these objections make this kind of experiment unlikely to take place today. Ethical concerns have been raised about the abuse allowed to be carried out on the prisoners, as well as over the mental stress suffered by all participants. There was a lack of protection to keep the subjects from physical or psychological harm. It has also been felt that the participants did not give fully informed consent before taking part in the experiment. They could not have consented in the full knowledge of what might happen as Zimbardo himself could not accurately predict the possible eventualities in advance.
Zimbardo's prison experiment remains influential today. This may be in part due to the fact that experiments of this kind are no longer possible with the ethical concerns and health and safety factors that come into play. Therefore there is currently little potential for any contradictory experimental evidence. However, the Stanford prison experiment still seems to have provided quite strong evidence for the factors discussed. It is an interesting example of how these societal behavioral influences appear to have such dramatic effects, especially in their seemingly excessive dominance over any behavioral responses arising from individual characteristics and personalities.