classic Milgram studies on obedience were inspired by the Nurnberg trials of Nazi war criminals who consistently argued in their defense to their charges that they were just carrying out orders. In his original study Stanley Milgram (1963) had wanted to see if people would inflict pain to the point of serious injury or death as the result of being ordered to do so by authority figures. Milgram used a sham learning experiment and a confederate learner to test his hypothesis that few people would actually progress to the point of inflicting damage on strangers at the bequest of an authority figure (as we will see his original hypothesis was incorrect). The learning experiment required a "learner" (the confederate) to learn a list of word pairs. The teachers (recruited participants) were required to administer what they believed to be were painful electrical shocks to the learner whenever the learner failed to recall a pair of words (the shocks were also a sham). The independent variable in this is the study is the authority figure and the prodding used to get the teacher to administer a shock to the learner. The dependent variable is the level of shock that the teacher would actually administer. The authority figure (Milgram in a white lab coat) would meet with both the teacher and learner before the start of the experiment. Unbeknownst to the teacher (a participant recruited by means of a local newspaper advertisement offering a payment of $4.00 to participate in a "learning experiment") the learner was always the same person (a confederate, who was part of the experiment). Milgram, dressed as an experimenter in a white laboratory jacket had the two people draw lots to see who would be the teacher and learner (of course this was fixed as mentioned above).
Once the experiment started the learner was taken to a different room and the teacher could not see him. They communicated via an intercom. The procedure of the experiment was rather simple. The teacher would read a list of word pairs to the learner. Once completed the teacher would go back and repeat the first word of the pair to the learner. The learner was supposed to reply with the second word of the pair. The teacher was seated or stood before a panel of knobs labeled according to the intensity of the shock that pressing each knob would deliver to the learner (the labels from 15 to 450 volts in 15 volt increments). If the learner got the second word in the pair incorrect, the teacher was told to administer a shock to him starting at 15 volts and increasing to the next level with each missed word. The shocks were a sham, and learner followed the same script for each participant, initially feigning being shocked with a grunt, later claiming that he had a bad heart and wanted to stop, and then not responding at all as if he had passed out.
When teachers became resistant to "shocking" the learners Milgram (the authority figure who was ordering the shocks; this is a case of expert authority as Milgram ran the experiment under the auspices of Yale University) used up to four verbal prods to get them to continue. The experiment was terminated after the fourth prod or after the participant had administered the 450 volt shock three times. Milgram had pre-tested the experiment with university students by explaining his proposed procedure to them. The students predicted that fewer than 10% would follow through to the ceiling level of shock of 450 volts and Milgram predicted that even fewer than that would go that far; however, during the experiment 65% of the subjects went to the 450 volt level, despite the learner's earlier protests, grunts of pain, complaints of a bad heart, and silence when high levels of painful shocks had been reached. Milgram followed his original experiment with several of versions (see Brown, 1986 for an excellent review of all of the Milgram studies).
This experiment was demonstrated to be quite reliable by follow-up studies. Milgram varied the distance between teacher and learner (which would be another independent variable, thus making the study a factorial design) and found that the tendency to obey his authority decreased the closer and more visible the learner was to the teacher. The initial study was considered so valid in its findings that it helped stimulate the formation of Institutional Review Boards in research universities and hospitals (see below). The study was also performed before the popularity of randomized controlled studies. By today's standards, the control condition would be considered inadequate (polling students as to how they would react in such an experiment) and the lack of randomized assignment would lead to some questioning of the cause and effect inferences that this study made. For example, using a control condition without an authority figure present to prod the teacher and randomly assigning participants to the control or experimental group (a condition with an authority figure) would allow for a more appropriate inference of authority causing the teacher to continue to shock the learner. Moreover, the sampling method by convenience or the recruiting method used in this study could to lead to some issues regarding generalization.
This experiment could not be performed today due to the potential for lasting psychological harm to be experienced by participants (the concept of minimal risk), participants were not informed that their participation could lead to later psychological distress (informed consent), Milgram's paradigm did not allow or inform participants that they had the right to withdraw at any time (informed consent), and the use of electrical shock, even shamed shocks, is now generally forbidden with human subjects by the (Sales & Folkman, 2000). None of these requirements existed when Milgram carried out his studies. The use of deception, as in the use of a confederate in this case, is not unethical, but must be justified by the paradigm and theoretical backdrop of the procedure. Moreover, debriefing in instances when deception is used is required.
Milgram's experiment actually led to the formation of stricter guidelines for the use of human subjects by the American Psychological Association. Milgram was investigated for ethical violations and was able to demonstrate that his participants did not generally suffer psychological harm as the majority later said that they believed they were better off after having performed the experiment and only one subject felt that they were worse off after completing the experiment (Milgram, 1974). This might be completed today using computer generated graphics or a virtual reality type presentation. However, strict adherence to getting informed consent and debriefing would need to be followed. The experiment could not be effectively performed by asking opinions or attitudes, as we have already seen Milgram and his students were way off in their prediction of how many participants would administer full level shocks.
Deception in psychological research and experimentation is often necessary. Deception involves giving subjects the wrong information or setting up sham conditions such as the use of confederates, it does not involve not telling subjects about the potential effects an experiment may have on them (Hertwig & Ortmann, 2008). Deception is often necessary because many participants will choose to follow their conceived notions of what a politically correct response should be if told what the experiment is actually measuring. For instance, a classic study in conformity by Solomon Asch (1951) could not have worked without deception. Asch, like Milgram, was interested in decision making under the pressure of bowing to others. He designed a now classic experiment to look at how people responded to the pressure to conform. Participants were recruited to estimate the length of lines. When they arrived they were seated at the end of a long table with several other students (groups of five or seven). The participants were shown two or three lines on a projection screen and asked if the lines were equal in length or differed in length. All of the students were confederates except the one at the end of the table. Many of the line pairs were obviously unequal in length, others equal. In a control condition all confederates gave the correct answer before the participant gave his/her response. In this condition only one response in hundreds given by the participant was incorrect. However, in the experimental conditions confederates verbally expressed that lines that were obviously unequal in length were equal. In these conditions three quarters of participants agreed with the obviously wrong answer at least once, and five percent agreed on every trial. Later, participants when questioned generally said that they did not believe the lines were equal in length but did not want to disagree with the group. Asch went on to design several other experiments to tests the limits of this tendency for many of us to conform to others using different groups sizes, agreement rates between confederates, etc. (Asch, 1956). Without the use of deception and confederates participants would have never acted in a…