Korva is a scroll of the internet’s best ideas.
Teens picked $10 over $11
A Split Made Up
In 1979, social psychologists Henri Tajfel and John Turner, working with Rupert Brown, tested how little it takes to make people favor their own side. They had 62 students, all 14 and 15 years old, look at abstract paintings and pick a favorite. The kids believed that choice put them on Team Triangle, Team Circle, Team Dichromatic, or Team Polychromatic. In fact they had been split arbitrarily into two groups, and the paintings had nothing to do with it.
Choosing the Pay
Next, the students filled out forms that set how much each group would be paid for the study, with a proposed amount for each group listed side by side. Some options were equal and some were not. Students kept choosing the unequal ones. Where both groups could have earned $11, they took $10 for their own group as long as the other group got $5.
What It Showed
The researchers read the choices as an effort to raise the status of their group. Making twice as much as the other group mattered more than making the most money. These groups shared nothing and had no conflict, which made the result evidence that bias can show up with no conflict at all.
Why build groups out of painting preferences?
Tajfel wanted the thinnest possible basis for a group, so he stripped out everything that normally divides people: shared history, contact, and any stake in the outcome. If bias still appeared, being sorted would be enough to cause it. Early versions of the experiment, run in Bristol in the early 1970s, used paintings by Paul Klee and Wassily Kandinsky. Participants never met the other members of either group, and they only knew each person by a code number.
How do payment matrices pit fairness against favoritism?
Each matrix pairs two numbers, one for each group, and a single choice sets both. Some options give the most to both groups combined, some give the most to the chooser's own group, and some maximize the gap between the groups. Tajfel's team found that students often gave up joint profit, and even their own group's profit, to widen that gap. Choices near the fair middle of the matrix also showed up throughout the data.
How well did the result hold up later?
Researchers repeated the minimal group method with many kinds of arbitrary labels. In 1992, Brendan Mullen, Rupert Brown, and Colleen Smith pooled a large set of these studies in a meta-analysis published in the European Journal of Social Psychology. They found reliable in-group bias across the experiments.
More from Stuff You Should Know
