Solomon Asch’s conformity experiments. · published 4 September 2026
The famous 75 per cent is the share of participants who caved at least once across the whole run, not the share of answers — counted by response, roughly two thirds of answers on critical trials were independent, about a quarter of people never conformed at all, and when Asch gave the participant one confederate telling the truth, conformity collapsed to a fraction of its level, which is why he wrote that what struck him was people’s independence.
These are the working notes, unedited. Everything the video
states as fact comes from here, and nothing else may be stated as fact.
The sections on limits and on what the finding does not show are
part of the brief, not an afterthought — see
the method.
Run at Swarthmore College and published from 1951, most fully in Psychological Monographs in 1956.
The setup: a participant sits with a group and is shown a standard line and three comparison lines. The task is trivial — which comparison line matches the standard. The answer is not ambiguous. In a control condition with nobody else present, people were right well over 99 per cent of the time.
Everyone else in the room is working for Asch. On certain trials they all give the same wrong answer, out loud, before the real participant answers.
The version everybody repeats is that people conform to a group even when the group is obviously wrong, and it is usually attached to a figure of about 75 per cent. That figure is real, and it is being used to say something the study did not find.
What the data actually show
Seventy-five per cent is the share of participants who went along with the group at least once across the whole run. It is not the share of answers that conformed. Counted by response, roughly a third of answers on the critical trials followed the group, which means about two thirds did not. The most common thing a participant did was give the right answer while the room gave the wrong one.
About a quarter of participants never conformed on a single trial.
The finding Asch thought mattered most is the one that almost never gets repeated. When he broke the unanimity — one confederate, giving the correct answer, against the rest — conformity collapsed to a fraction of its previous level, in the region of a quarter of it. The exact figures vary between conditions and should be given as an order of magnitude rather than a precise number. It was not the size of the majority that mattered so much as whether it was unanimous: a single visible ally restored independence.
Asch wrote about being struck by the independence people showed, not by the conformity. The popular telling reverses his emphasis.
Limits
The original participants were all male college students, and the samples were small. Bond and Smith’s meta-analysis (1996) pooled 133 studies across 17 countries and found the effect is real but varies by culture, and that measured conformity in United States samples declined across the decades after Asch ran it. It is not a fixed constant of human nature.