Abstract
The mental model theory of reasoning postulates that individuals reason by constructing models of the situations described by premises, but the number of explicit models and the information in them is kept to a minimum. Among the novel predictions made by the theory are the following three: I reasoning with conditionals should be easier than reasoning with exclusive disjunctions; 2. reasoning with a minor premise that affirms a constituent of a major premise should be easier than reasoning with a minor premise that negates a constituent of a major premise; and 3. these two variables should interact, because reasoning with a premise that negates a constituent of a conditional calls for the conditional to be fleshed out to two or three explicit models. The present paper reports an experiment that corroborates all three predictions.