Correlate, not cause
One of the factors human minds sometimes are misled by is correlates that are not causes. Let’s say that we find that pickpockets are more often ambidextrous than the general population. So: does favoring the use of both hands equally lead one into pickpocketry?
That idea might be the focus of research for a while. I am not trying to start a rumor about ambidextrians. The internet says that genuine ambidextry occurs in about 1% of people.