Becoming a teacher can reduce obedience compared to being solely an examiner. Agentic state and obedience in the Milgram paradigm.

Studies of obedience carried out in the Milgram paradigm tend to report shockingly high levels of obedience from people who are ordered by an authority figure to eventually, if administer all required shocks, electrocute another person. In the psychology literature, the person who carries out these commands is called the teacher. The authors of the present article note, however, that the term "examiner" would be more appropriate here, since the study participant is limited to verifying the correctness of the responses given by the student, i.e., the person sitting behind the wall. It was assumed that if the participant actually performed the role of a teacher (and thus first taught the "student," and only then checked the correctness of the answers to questions), the level of obedience demonstrated would be reduced. The results of our experiment partially confirmed this assumption. In the examiner condition, 4 out of 40 participants (10%) refused to press all ten switches, meaning that 90% proceeded to 150 V. In the conditions where participants had first taught the student, refusals occurred more than twice as often: 9 out of 40 (22.5%), with 77.5% reaching 150 V. This difference, however, was not statistically significant. We also analyzed an indirect measure of non-compliance-the frequency with which the experimenter had to prompt participants to continue by reciting the standardized phrases prescribed by the procedure whenever participants expressed hesitation or refused to comply. These experimenter interventions were more frequent in the conditions where participants had previously taught the learner (Median = 0.5) than in the examiner condition, where their role was limited to punishing learner for his mistakes (Median = 0). This difference was statistically significant.