Deductive Reasoning

Sources of Evidence

The psychological study of deductive reasoning has largely focused on investigating its algorithmic underpinnings. Experimental evidence has come from the performance of subjects, typically undergraduate students, on specific reasoning tasks. In one experimental paradigm, subjects are presented with premises and asked to indicate whether a proposed conclusion follows. In a related paradigm, subjects are presented with premises and asked to generate a conclusion that follows, if one exists.

A notable finding of this research is that subjects are highly prone to errors in evaluating the validity of arguments. For example, a meta-analysis carried out by Walter Schroyens, Walter Schaeken, and Géry D'Ydewalle of 65 studies on conditional reasoning found that while 97% of subjects correctly evaluated modus ponens inferences (if p, then q; p, therefore q) as valid, only 72% correctly evaluated modus tollens inferences (if p, then q; not-q, therefore not-p) as valid. Worse still, 63% incorrectly identified instances of the fallacy of affirming the consequent (if p, then q; q, therefore p) as valid and 55% incorrectly identified instances of the fallacy of denying the antecedent (if p, then q; not-p, therefore not-q) as valid. This research has also uncovered several content effects that influence subjects' responses. For example, subjects are more likely to mistakenly identify an invalid argument as valid if its conclusion is believable.

...