Just Story

The Latest and Most Important News

We need to radically change the way science works

2 min read

Forum. The scientific study of human behavior provides relevant and useful knowledge for solving societal problems are urgent and complex, such as depression, discrimination and climate change. Prior to 2011, many researchers thought that the scientific process was effective. We were completely in error. More importantly, our field has discovered that even of the honest searchers could produce knowledge that is non-reliable. Thus, we need to radically change the way science works.

If so is it spent in 2011 ? Daryl Bem (Cornell university, Usa), a well-known psychologist, has found evidence that in appearance very strong in the human being’s ability to see in the future using the methodological criteria shared by many disciplines. Subsequent studies have demonstrated that this assertion implausible was false.

In addition, a member of our collective has conducted two international projects of replication of 51 other original studies, in which assertions were tested on more than 15 000 participants. In all, only 27 of the 51 statements previously reported in the scientific literature have been confirmed. These problems are not limited to psychology, they have been detected in many disciplines (such as biology of cancer). Without verification of their work with methods that are routine, researchers can thus find evidence in data that in reality do not contain. This should be a concern for all, since the knowledge of the behavioral sciences are sociétalement important.

Failures must also be known

But what are the problems ? First, it is difficult to verify the integrity of the data and equipment used because they are not shared freely and openly. When researchers have requested data from 141 articles published in major journals in psychology, they have received that in 27% of cases. In addition, errors were more frequent in the the items for which data were not available. Then, a publication of Joe Simmons and his colleagues has demonstrated how it is possible to use statistics to prove any scientific idea.

At the same time, most of the time, we do not have knowledge of the failures of scientific or even of a priori hypotheses of the researchers. In the majority of scientific fields, only the success of the researchers are published and their failures go in the trash. Imagine that it happens the same way with the sport : if the Olympique de Marseille failed to communicate that his victories and hidden defeats, one might think (wrongly) that this is an excellent team. We do not tolerate this approach in the sports field. Why should we tolerate it in the field ?