In the 1960s, Stanley Milgrim, then a psychologist at Yale University, conducted a controversial experiment. In an obviously controlled setting for the study, Milgrim and his associates directed their subjects to give ever-increasing electrical shocks to strangers whenever they gave the wrong answer to questions in a test of memory. The strangers were really actors pretending to experience pain and did not actually receive any jolts of electricity. The study, intended to measure how compliant people would be to obey authority figures, even to the point of inflicting pain on otherwise innocent individuals, was disturbing, to say the least. The subjects, despite some discomfort at first, continued to “shock” the test-taking actors. The experiment raised ethical concerns about subject study methodologies, among other things.
However, recently, Mel Slater, a computer scientist at University College London in England (with a joint appointment to the Universitat Politecnica de Catalunya in Barcelona), reproduced this experiment without running afoul of the ethical concerns that the original study raised. Mr. Slater conducted his experiment in a virtual world where test-taking “victims” were avatars whose expressions changed from happy to pained in response to the actions of participants in the study—much the same way the real-life actors did more than 40 years ago. Not only were the results astonishingly similar, but to the surprise of many, the real-life participants experienced increased heart rates and described feelings of regret or feeling badly about delivering electrical shocks—even though they knew the strangers on the other side of the screens were avatars and not real people!