When cooperating with a partner, what attitude do you adopt? In an experiment carried out by researchers from the Carlos III University of Madrid (Spain) together with colleagues from the universities of Barcelona, Rovira i Virgili and Zaragoza with 541 volunteers, 90% of the participants fit into four character classes: pessimistic , optimistic, envious or confident.
The experimental study posed 100 social dilemmas for volunteers to observe their response to peers. Would they enter into an internal conflict or would they directly betray him? The experiment – within game theory – posed problems involving both individual and collective interests.
What behavior do we adopt when faced with a dilemma? And how do others react?
“These people were asked to play in pairs that changed in each round, but also each time the game was different , that is, the best thing could be to cooperate with the other, do the opposite of the other or betray him. In this way we have data on what they do in very different social situations ”, explains one of the study authors, Anxo Sánchez, from the Carlos III University of Madrid (UC3M).
Following the social experiment, the researchers developed a computer algorithm to classify the participants based on their behavior. The algorithm grouped 90% into four groups: the envious (they do not care how much they earn as long as it is more than the others) represented 30% of the participants; the optimists (who decide thinking that the other will choose the best for both) were 20%; the pessimists (the least bad option) were 20%; and, finally, the confident ones (they always cooperate and don’t care about winning or losing) were 20% of the volunteers.
The remaining 10% is made up of a fifth group that the computer algorithm was unable to classify because they did not fit into any of the other four personality patterns.
“The curious thing is that the classification was made by a computer algorithm that could have obtained a large number of groups and, nevertheless, has yielded an optimal classification in four types of characters,” explains Yamir Moreno, co-author of the work.
“Previously, the experiments were carried out with dozens of people. With this platform, the volume of study participants can be increased , as well as tests with a heterogeneous population and record much more data on how they act. This has opened the door to much more complex tests ”, says Jordi Duch, also co-author of the study that has been published in the journal Science Advances.
