Today we will talk about judgments and decisions that are made when other people are present. These people aren’t explicitly trying to change our behavior, but often their mere presence is enough to cause us to do something else. Most people have a drive to copy what the people around them are doing. That is, most people have a desire to conform.
Conformity was first studied in the lab by Muzafer Sherif (pictured below) in the 1930s.
In Sherif’s experiment, he put people in a dark room with a single point of light on one of the walls. This looks something like this, with the difference that the whole rest of the room was also pitch black.
In this situation the light appears to move on its own due to involuntary eye movements (microsaccades). This effect is called the Autokinetic Effect.
What made this a social experiment — and not just an experiment in perception — was that people made their judgment in groups. Moreover, participants came back to repeat the experiment with the same group of people on multiple days.
On the first day, different people reported different directions. But by the second day, they all reported the same direction, even though the direction of motion in the Autokinetic Effect is largely random. That is, people conformed with the judgment of the group and the changes in judgment were used as a model of culture formation.
Another early experiment on conformity was by Solomon Asch in 1956. In this study people had to make judgments about the lengths of lines. In particular, they had to decide which of the three lines on the right best matched the example line on the left.
When people make this judgment alone they are close to perfect, making errors less than 1% of the time.
However, people perform much worse when, before reporting their own judgment, they see 7 other people (who were actors) make the wrong judgment first. In this case people made conformed with the others, and hence made obvious mistakes, about 33% of the time.
This study has been replicated multiple times around the world, with a meta analysis from 1996 citing 133 studies. This meta analysis showed that
So the pessimistic conclusion from this work is that we are all Sheeple, doomed to conform to make bad judgments even when the correct answer is obvious
Hodges and Geyer 2006 took a second look at data from the Asch experiment. Even in the original study they found that conformity isn’t actually that high. In fact, most people only conformed a few times out of 12 trials. Here’s the distribution over the number of times people conform.
In addition, they noted that Asch also interviewed his participants on their motivates for conforming. It turned out that everyone knew the correct answer, but most people were just trying to get along with people!
So, yes people do conform, but they do it for the good reason of being pro-social!
Of course, measuring an effect in the lab doesn’t mean that that effect is relavent for real world behavior. To test whether conformity could be used to change behavior in the real world, Coultas 2004 performed an experiment in a mid-2000s computer lab. In case you’re unfamiliar with what these looked like, here’s an archive picture …
This study asked whether people would adopt a new behavior if they saw others doing it. The behavior was all about the keyboard covers on the computer and what to do with them. Keyboard covers were pieces of plastic that went over the keys to prevent them from gathering dust when not in use. They looked like this
The behavior they wanted people to adopt was placing the keyboard cover on the top of their computers. This was not something people usually did, but the experimenters wondered that if they had actors in the room who did this, whether other people would copy them …
The manipulation was the fraction of actors in the lab. Which ranged from 0 to close to 1 (e.g. all but one person is an actor). They found that people were more likely to conform the more actors were in the room.
This study shows that conformity could occur in the real world. While not so carefully tested, other examples suggest that conformity occurs in the real world. In much of the early work on conformity, the focus was on how conformity can be bad …
Conformity can definitely be bad in some cases. For example, with smoking. Tobacco companies spent a lot of money encouraging movies to include scenes with people smoking. The whole point of this was to make (and keep) smoking as a socially acceptable activity that everyone was doing.
Another example is hooliganism in soccer. Sadly this is something that my own country (England) is famous for. Why do people act so badly over something so trivial as a soccer match? In part because “everyone else is doing it,” they are conforming with the culture.
Early research often framed conformity as a bad thing. In some ways, by doing so, the authors were themselves conforming to the dominant culture of the times.
In particular, this early work was done in the 50s and 60s, shorly after World War 2 — where conformity with nazi ideology caused the deaths of millions — and in the middle of the Cold War against communism — a system of government in which conformity is prized. In addition, for the later work it was the 60s, a time when there was widespread pushback against the conformity that had defined earlier times.
But while conformity can definitely be bad, in many cases can also be good. We know from the studies on the Wisdom of the Crowd that groups of people can make good decisions. In that case, it may be wise to copy them!
Conformity can also be good because it allows us to agree on things. For example, the meaning of the word “green.”
What we call “colors” are really just photons of light of different wavelengths reaching our eyes. The wavelength of the light is actually a continuous number, but we group ranges these wavelengths into groups and call them colors.
For these names to make sense we all have to agree on their definition, that is we have to conform to the culture and conformity is useful in this case because it allows us to communicate with each other.
One example is green. In European languages this covers the range 520-570nm (nm stands for nanometers, the wavelengths of visible light are very small). So (as a native English speaker, which is a European language) when I tell you something is green you understand what I mean.