Belief bias in conditional reasoning: Exploring the role of content in the mental models theory

This study examined whether the popular mental models theory is a viable account of belief bias in conditional reasoning. Although in focusing on representation the mental models theory would appear to facilitate the exploration of both problem content and belief, its proponents typically dismiss the former factor as irrelevant to reasoning. Since recent trends support a more pragmatic view of reasoning that appreciates the wide range of content encountered in daily life, it seems that a revision of the mental models theory is warranted. The present study serves as an attempt to update this approach by unpacking the concept of belief through direct investigation of non-logical factors such as problem content.; This study was designed to highlight the role of problem content in belief bias by manipulating the truthfulness of premise statements and the believability of conclusions across two content conditions (abstract-pictorial and thematic-verbal). Participants were one hundred college age volunteers recruited from undergraduate psychology classes. In modifying the work of Cherubini, Garnham, Oakhill, & Morley (1998), an eight item test of conditional reasoning was developed by repeating the two valid problem forms and the two logical fallacies across abstract and thematic content conditions.; The findings identified some significant effects for premise truthfulness, the believability of conclusions, and problem content, although not across all problem types. Patterns of response suggest that belief bias and the effects of content operate differently depending upon the difficulty level of a particular problem. In addition, there were some interesting three-way interactions between truthfulness, believability, and problem type. The effects were found for the more difficult fallacies; here it seems that a discrepancy between the truthfulness of the premise and believability of the conclusion was conducive to both valid and logical fallacy responses.; The findings support a revision of the mental models theory to include problem content as an important factor in conditional reasoning. Further, the results underscore the utility of including the more difficult logical fallacies in belief bias study, as gradations in problem difficulty appear to interact with belief bias and content effects in meaningful, identifiable ways.