A handful of psychology experiments are famous enough to turn up in films, ads and dinner-party arguments. They're worth knowing, but it's just as important to know what they actually showed, because several of them are more complicated than the usual retelling.
What the classics claimed
- Obedience (Milgram). Ordinary people kept giving what they believed were dangerous shocks when an authority figure told them to. The lesson people usually draw: situations can push us further than we'd ever expect of ourselves.
- Conformity (Asch). Most people went along with an obviously wrong group answer at least once. But a single person who disagreed with the group made it much easier to resist.
- Learning (Pavlov and Bandura). We learn by association, linking a cue with a result, and by watching what others do.
- Helping (the bystander effect). The more people who are present, the less likely any one of them is to step in.
Why some of them have wobbled
From the early 2010s, researchers began repeating well-known studies with bigger samples and their methods fixed in advance. In one large 2015 effort, only about a third of 100 published results came out the same way again. This became known as the replication crisis.
Some famous examples look different now:
- The Stanford prison experiment is often told as "give people a uniform and they turn cruel". But the guards were coached toward harshness, so it shows much less than people assume.
- The marshmallow test link between waiting as a child and success later shrank a lot in a larger study that took family background into account.
- The Kitty Genovese story, which started bystander research, was exaggerated by the press. The bystander effect itself has still been found in later work.
How to read a psychology headline
Ask three things: how many people were studied, has anyone repeated it, and is the claim bigger than what was measured? A study isn't wrong just because it's old or surprising, but one result is a lead, not a law.
Try it
The cards below cover each study, plus a few on how the field fixes its own mistakes.