photo from: nplus1.ru
Scientists were able to reproduce the results of one of the most influential works in the field of behavioral Economics.
Scientists were able to reproduce the results of one of the most influential works in the field of behavioral Economics — the original article about the theory of perspectives, written by Daniel Kahneman and Amos Tversky in 1979. To reproduce, the scientists repeatedly increased the sample, while also adapted the original surveys for 19 countries. On average, were able to reproduce 94% of the results and to confirm 12 of the 13 examined in the original work of the manifestations of the behavioral effects, writes nplus1.ru with reference to the Nature of Human Behaviour.
One of the most influential theories in behavioral Economics, prospect theory argues that when making decisions in a situation of possible risk the person relies on its own subjective judgments about the usefulness in the evaluation of winnings and losses and never behaves rationally — contrary to the predictions of classical economic theory. Detailed prospect theory described in the article Daniel Kahneman and Amos Tversky published in 1979: it is related to the theory effects (e.g. effect of confidence) scientists also tested experimentally.
Despite the fact that prospect theory since the early 1980s, not only widely used in psychology and behavioral Economics, but takes into account classic economic theory (in 2002, Kahneman even won the Nobel prize in Economics) in the original study still has some limitations. Each of 20 tasks, divided into 13 oppositions election that scientists have developed to test his theory, decided less than a hundred people, besides they were all students (though from different countries: Israel, USA and Sweden).
Of course, the theory of perspectives, not just the experimentally tested in numerous studies, but the playback results are the original work of Kahneman and Tversky is still almost no one did — especially on a bigger sample. To do this it was decided by scientists under the leadership of Thomas Folke (Tomas Folke) from Columbia University: they adapted the tasks from the original work of Kahneman and Tversky for 4098 inhabitants 19 countries and translated them into 13 languages respondents. The number of jobs was reduced from 20 to 17, but they also covered all 13 of the assumptions.
For example, to check the effect of confidence, according to which option one hundred percent probability probability preferable to the other options, the researchers used the following example from the original work:
In the experiment of Kahneman and Tversky 82 percent of respondents voted for the second option, thus confirming, on the one hand, the effect of confidence: despite the fact that in the first option the possibility of not getting anything is minimal, and also quite a high probability to get more, people still choose the safe option with absolute certainty.
Replication of these results was considered successful if the direction of the effect (a preference for a certain option) was the same as in the original study, and differed from the level of accidental exposure. Scientists were able to reproduce 94% of the results that were obtained by Kahneman and Tversky: in some countries (mainland China, Hong Kong and Australia), the reproducibility reached 100 percent.
Could not reproduce only the choice of one contrasting, reflective effect of the mirror: in this opposition included two options (receive $ 4,000 with probability 20% and 3000 with a probability of 25 percent, or lose the same amount with the same probability). In the original work the majority of participants chose more winning with a lower probability and lower loss are more likely, whereas in the new study — on the contrary. However, scientists noted that other options, checking the mirror effect, to play was able, as in the original study, the distribution of answers when you select one of these options was not much above the accidental.
Scientists thus were able to reproduce experimentally prove various aspects of prospect theory. Interestingly, a survey conducted after the experiments, indicated that about one third of the participants were familiar with some concepts of the theory — but the results is not affected.