| In traditional tasks of formal reasoning, participants are asked to evaluate the validity of logical arguments. While this research tradition has contributed in many ways to our understanding of human reasoning, the extent to which this body of research generalizes to everyday, or informal, reasoning is unclear (e.g., Evans & Thompson, 2004; Galotti, 1989). The main goal of this dissertation was to illustrate the benefits of applying an informal approach to the study of conditional reasoning. In six experiments, everyday conditionals in the form of inducements (promises and threats) and advice (tips and warnings) were investigated. The results support three main conclusions. First, people recruit a substantial amount of background knowledge when interpreting and reasoning with these conditionals. Specifically, inducements were found to be different from advice on several pragmatic variables (Experiment 1); these variables also predicted differences in inference patterns (Experiment 2). Second, these studies provide further support for a probabilistic interpretation of conditionals (e.g., Evans & Over, 2004; Oaksford & Chater, 2001). Thus, in Experiments 3-5, estimates of different conditional probabilities predicted a number of judgments people make about inducements and advice. A particularly interesting finding was that the effectiveness of these conditionals in changing behaviour did not seem to depend on how likely they were perceived to be true. Finally, Experiment 6 adopted a decision-theoretic analysis (e.g., Over, Manktelow, & Hadjichristidis, 2004), showing that the effectiveness and quality of inducements and advice were tied to perceptions of subjective utility and preferences among possible outcomes. This dissertation concludes with a theoretical discussion of the nature of the relationship between formal and informal reasoning. |
