Investigating Theories of Belief Bias: Individual Differences in Interpretation, Strategy Selection and Performance

Belief bias, defined as the tendency to endorse conclusions that are consistent with one’s beliefs, rather than on the basis of logic, is one of the clearest demonstrations of the variable nature of deductive inference. Indeed, the interactions between belief and validity form the foundation of many dual process theories of reasoning and belief bias and the investigation of the effects of belief bias has been a primary test of such dual-process theories. The dual strategy model, proposed by Verschueren, Schaeken and d’Ydawelle (2005a; 2005b) has been proposed as an explanation of the variability of reasoning. The model suggests that individuals have access to two qualitatively different strategies with which to approach deductive reasoning, statistical and counterexample. Markovits and colleagues have provided strong evidence for the distinction between these strategies. This PhD comprised of four experiments, which investigated theories of the biasing effect of belief on reasoning by examining: individual differences in interpretation, strategy selection and performance. The first two studies extended the work of Markovits and colleagues by using syllogistic reasoning problems, and triangulating measures of responses, response times, and individual differences in disposition and interpretation. Study Three adopted an eye-tracking methodology, to further investigate these individual differences and explore the metacognitive elements of reasoning in more depth. The final study then examined individual differences in interpretation, strategy selection and performance when reasoning about information related to personally held beliefs. The findings of these studies were broadly consistent with the predictions of the dual strategy model, demonstrating differences between statistical and counterexample reasoners. Moreover, the results provided evidence that the preferred strategies adopted by reasoners on reasoning tasks are malleable based on the instruction set provide. The effects observed around figure also offer some novel insight into the differences between strategies and the way in which information is processed and quantifier interpretation was also found to play an important role. Results relating to metacognitive indices were also consistent with reasoners having access to qualitatively different strategies. Novel insights into to the effects of personally held beliefs are offered, providing further evidence for the dual-strategy model. Finally, this programme of research highlighted the importance of the triangulation of not only performance measures (acceptance rates, response times and confidence ratings), but also individual difference measures such as thinking disposition, interpretation and belief, in addition to the consideration of metacognition and behavioural measures (such as eye-tracking) in the development of a complete theory of reasoning.