ABSTRACT
Braine's theory of mental predicate logic (unpublished ms.) will be briefly introduced. A preliminary empirical examination for a major subset of 10 schemas proposed by Braine's model will be reported. The methodology is a modified version of that used by Braine, Reiser, & Rumain (1984). We design a large number (64) of reasoning problems of predicate logic sort-most fairly simple-for which the model predicts the reasoning that subjects will use to solve them. In each problem, one or more facts are given-these serve as premises, followed by a conclusion to be evaluated. Subjects must decide whether the proposed conclusion is true or false, given those facts. They are also to rate each problem's subjective difficulty on a 7-point scale right after evaluating the conclusion given. Within this set of problems, the number of reasoning steps is varied. The mental logic must account (among other things) for which reasoning steps are easy and which are difficult, and it is expected that adding additional steps will make a problem more difficult.
