However, Milgram’s use of a laboratory experiment means that the findings and conclusions drawn from it lack external validity. Research took place in an artificial and controlled environment and therefore wouldn’t be reflective of obedience in real life. This means that the findings and conclusions about obedience are situation specific and cannot be generalised beyond their artificial setting.
Orne criticised the study for producing demand characteristics. Because a laboratory experiment was used, it is possible that the participants received cues as to the type of behaviour expected from them, and they could have guessed the aims of the research, knowing that the shocks weren’t real, which would have flawed the effects of an authority figure. If this is the case then the study’s internal validity is threatened.
Milgram carried out a number of variations to his original procedure. The physical and therefore emotional distance of the teacher from the learner changed levels of obedience. When the participants had to force the learners hands down onto the shock plates, obedience levels dropped from 65% in the original study, to 30%. When the experimenter gave orders to the teacher down the phone, obedience levels dropped even further to 20%. Furthermore, when the experiment was taken to a run down office block, so that it wasn’t related to Yale University, obedience levels dropped to approximately 40%.
The findings show that location is important in determining obedience, but factors such as the presence of a perceived authority figure are more important.
The variations are perhaps more representative of obedience in real life, for example, in Nazi Germany, soldiers were often obeying orders from an officer who wasn’t present at the time.
Bickman carried out research which found that obedience was more likely when the authority figure was dressed in particular clothing. This would be a lab coat in Milgram’s study, and a guard’s uniform in Bickman’s. Due to it being a field experiment, it has good external validity, which suggests that Milgram’s findings are also representative of obedience in real life situations as they both found similar results.
Orne criticised Milgram’s study for producing demand characteristics. He claims that participants knew that the shocks weren’t real, and were therefore willing to administer the shocks without actually worrying about it.
Sheridan and King carried out research that contradicts Orne, and supports the findings of Milgram.
They found that 100% of females and 54% of males were willing to administer the full 450 volts to a puppy that they could clearly see. The puppy was an authentic victim, and as it produced similar results to Milgram, it can be argued that Milgram’s research wasn’t flawed by demand characteristics.
Critics argue that the study lacks external validity due to the controlled nature of experiments. Hofling carried out a study in a real life hospital situation and found that 21 out of 22 nurses were willing to administer a potentially lethal dose of a drug called Astroten to patients, just because they had been told to do so by a bogus doctor over the phone. Both studies can be given credit for bringing to our attention the dangers of unquestionable blind obedience.