Thus, the requirement of reproducibility meaningful only for those cases where the facts raise doubts themselves. It is particularly important when the fact is detected only by rather complex data transformations, for example by statistical processing. Psychologists empiric, unfortunately, very rarely checked, as, for example, the correlation found them in one study, replicated in the other. But, apparently, they realize that such testing is likely to lead them to a frustrating results.  So many of them are doing their best to forget about the fact that even if one day found a correlation is statistically significant, it only gives the right not to discard the hypothesis of the existence of a connection, but does not prove the existence of this communication.