Extending the Theory of Reasoning by Mental Models: Tests of New Predictions

Abstract

The theory of reasoning by mental models provides a far-reaching account of deductive reasoning processes covering syllogistic reasoning, propositional reasoning, spatial and temporal reasoning, among others. It is a semantic theory based on representational assumptions about mental models, a set of rules to manipulate such models, and assumptions about working memory involvement. The chapter reviews old and new data that address these fundamental assumptions in the domain of propositional reasoning. In particular, premise-reading time data are used to examine the representational assumptions about initial models, data from multinomial-modeling analyses test the predictions for problem difficulty, and data from dual-task paradigms address the working memory assumptions. Training studies finally provide insight into the interplay of different reasoning strategies with simplified fallback heuristics employed by reasoners working under cognitive load.