In of 1961, participants were asked to deliver what they believed were rising electric shocks to another person, on the instructions of a researcher with authority. The result has been quoted ever since as a single number summing up human nature. What is that number hiding?
Twenty-three other . Milgram ran twenty-four versions, each changing one detail, and obedience moved enormously. It fell when the victim sat in the same room, and further when the participant had to press the victim’s hand to a plate. It collapsed when the researcher gave instructions by telephone, and nearly vanished when two researchers disagreed with each other. So the real result is not a number but a map.
The figure everyone knows is that about two thirds of participants went to the end of the scale. It is correct, and it comes from one condition out of twenty-four.
Milgram himself ran many variations, each changing a single detail. When the victim sat in the same room, obedience fell. When the participant had to press the victim’s hand to a plate himself, it fell further. When the researcher gave instructions by telephone instead of sitting alongside, it collapsed. When two researchers disagreed with each other, almost nobody continued.
Which is to say the real result is not a number but a map. Obedience is acutely sensitive to distance from the victim, to the physical presence of the authority, and to whether that authority is united. That is an entirely different lesson from "two thirds of people would have carried on".
Modern replications, stopping earlier for reasons of , show similar patterns today. of the original data runs differently again: the participants were not blindly obedient but cooperating with what looked to them like an important scientific purpose. They were persuaded by "the experiment requires that you continue" rather than by threat, and that changes the conclusion too.
wat. Most of what people know about Milgram comes from one condition out of twenty-four.
wat wat
A more accurate picture matters because it changes the conclusions drawn about obedience to authority, rather than resting on one dramatic figure that is easy to quote and hard to learn from.
How do we know?
Ethical replications in recent decades repeat the conditions up to an earlier stopping point, and compare the obedience rates against Milgram’s own figures at that same point.
What is still unknown
An alternative reading of the original data argues the participants were not blindly obedient but cooperating with a goal that looked scientific and important to them.
Learn something?
Where to next
What happened when the Dunning-Kruger effect was run on random data?replication
What was left of the marshmallow test when they ran it on ten times as many children?replication
Does the full moon really keep people awake?replication