There is not a lot of empirical evidence for this, and even less evidence of any other kind.

What -does- have a lot of that, is the statement that humans are particularly bad at classifying random events.