The linked study compares people based on "numeracy," i.e. a facility with numerical data. It finds that people who score higher on this are better able to draw causal inferences from data when faced with a difficult question.
The problem, though, is that this advantage went out the window when it was something political. Once it become a hot-button topic (in this case evaluating a measure on gun control):
The authors conclude that their findings support the Identity-Protective Cognition Thesis, which says that cultural conflict overrides our reasoning abilities.
Our mental faculties evolved under conditions that selected for fitness. Apprehending reality in some sort of "objective" way is not necessarily a better adaptation.
But this is a well written paper. It's a nice alternative to the poorly written social science I've been subject to so far this semester.