Milgram found that no participants stopped the experiment before 300 V, and 65% actually continued as far as 450 V. Milgram concluded that ordinary Americans obey orders even if it leads to them acting against their conscience and hurting others.
This was a very influential study, which provided great insights into human behaviour. It disproved the `Germans are different` hypothesis, and led to increased awareness of how easily we can just blindly obey orders, without questioning whether we morally should. However, there are a number of criticisms, including ethical issues and issues of validity.
The participants in Milgram’s study suffered a lot of psychological distress. They were also deceived as to the nature of the experiment, which meant they couldn’t provide informed consent. Additionally, they were not informed of their right to withdraw, which is now common practice in psychology experiments. Instead, they were urged to continue with the experiment when they asked to stop. This experiment was therefore very ethically questionable, and would never be allowed today. However, Milgram extensively debriefed his participants, including reuniting them with the learner. The participants therefore left understanding that they hadn’t hurt anyone. They did, however, leave in the knowledge that they were capable of hurting people, which may have caused them distress.
Milgram’s study can also be criticised in terms of external and internal validity. Some people claim that it lacks internal validity – that it wasn’t actually measuring obedience rates at all. Perhaps instead, participants were just acting along with the experimenter, not actually believing they were hurting the learner (showing demand characteristics). The experiment has also been criticised in terms of external validity – that it doesn’t represent traits that would happen in the real world, with different people, and different situations. However, similar obedience rates have since been shown in other studies (e.g. Hofling et al), which do have high external validity.