Individual differences in reasoning Ines Skelac Faculty of humanities and social sciences, University of Rijeka Sveučilišna avenija 4, Rijeka, Croatia +385958368526 ines.skelac@gmail.com ABSTRACT The Mental Models Theory and The Mental Logic Theory, two famous theories of human reasoning, are commonly perceived as disjunctive theories. In this work we try to examine possibilities of their coexistence with respect of individual differences between reasoners, incorporating simplified versions of existing theories as possible strategies of reasoning. Spatial representation of information corresponds to the Mental Model Theory, while verbal (propositional) strategy can be connected to the Mental Logic Theory. Categories and Subject Descriptors The ACM Computing Classification Scheme is not fully applicable for this submission, but it can be partly connected with I.2 [Artificial Intelligence] I.2.0 General - Philosophical foundations; I.2.3 Deduction and Theorem Proving - Deduction (e.g., natural, rule-based); I.2.8 Problem Solving, Control Methods, and Search - Heuristic methods. General Terms Measurement, Theory Keywords Reasoning, individual differences, mental models, mental logic. 1. INTRODUCTION Individuals with no previous knowledge in formal logic are able to make deductive inferences. For example, for given arguments: If I have 1 euro in my pocket, I can buy a coffee. I have 1 euro in my pocket. they draw the conclusion: I can by a coffee. How individuals untrained in logic are able to draw valid conclusions or determine for given conclusion if it is valid is a matter of controversy. (Johnson-Laird, 2008). According to a very old idea, which origin is probably in the work of Aristotle, logic is concerned with discovering or illuminating the laws of thought. Its psychological corollary is that a system of logic (but maybe not the same as the first order logic) in the mind underlines our thinking processes. It is called The Mental Logic Theory (ML) and it holds that deduction rules are the exclusive tools of thought and abstract rules are applied to verbal/propositional representations. Evidence for this theory appeared mostly in the last few decades (see Rips, 1983). In the same time a new theory has been formulated: The Mental Models Theory (MM), mostly due to the work of Johnson-Laird and his collaborators. MM also deals with the nature of the internal representations of deductive processes, but these theories differ in their supposed nature. The MM hypothesis posits that the engine of human reasoning relies on content, so valid inferences can be drawn without explicit representations of logical properties of relations. Instead, information is represented in the form of spatial arrays, akin to mental diagrams, from which further information can be inferred (see Johnson-Laird & Byrne, 1991). To infer that the inference is valid, a reasoner should try to find a counterexample in which premises are true, but the conclusion is false. If there is no such counterexample, the inference is valid (Johnson-Laird, 2008). ML and MM are commonly perceived as disjunctive theories. In this work we will try to examine possibilities of their coexistence with respect of individual differences between reasoners, incorporating simplified versions of existing theories as possible strategies of reasoning. Recent researches have suggested individual differences in strategies that reasoners use in solving reasoning (syllogistic) tasks and this is connected with other cognitive processes. Bacon, Handley, and Newstead (2003) and Ford (1995) have suggested that while some people prefer to use spatial representation of information, others are more likely to use verbal (propositional) strategy. Furthermore, Bacon et al (2008) connected strategy that people choose when reasoning with better performance of verbal of spatial working memory. 2. REASONING Reasoning is a fundamental cognitive activity which is present in every situation in which one needs to reach some conclusion. It involves the manipulation and the transformation (i.e. coding) of information in order to make inferences. A lot of researches that examine the process of reasoning use tasks including syllogisms, logical arguments containing two premises and conclusion. Conclusion can logically follow or not follow from premises and one of the most frequent tasks given to participants in those 37