.png){kind=icon}
Chapter Summary
This chapter examines how we think when we are reasoning or making judgments and choices. Reasoning refers to the process of thought by which conclusions are drawn “from percepts, thoughts, or assertions” (collectively referred to as premises). The oldest form of reasoning dates back to Aristotle and is called syllogistic reasoning (or categorical reasoning) which involves drawing a conclusion from two given premises. Each premise can belong to one of the following categories: universal affirmative, universal negative, particular affirmative, or particular negative. It is possible to interpret the premises of a syllogism in many ways, which can lead to complications when judging the soundness of a syllogism, especially when considering that a valid syllogism requires that the conclusion follow from the premises, regardless of whether or not the premises are correct. If the conclusion fails to follow logically from the premises, the syllogism is invalid.
Logicism refers to the belief that logical reasoning is an indispensable aspect of human nature, which is often illustrated through practical syllogism. However, untrained people often make logical errors in everyday life, indicating that processes other than logic affect how conclusions are drawn. Studies show that the subject area in which a person is reasoning exerts an influence on the reasoning process, as do beliefs and experiences. Furthermore, the logical meaning of the word “some” is quite different from its popular meaning, and this difference can also introduce errors of logical deduction. In logic, some means at least one, and possibly all.
According to Johnson-Laird’s theory of syllogistic reasoning, people use mental models to represent a situation and to draw conclusions about it. Mental model theories have expanded and now include concepts such as relational reasoning (premises that include relations between items, e.g., A is taller than B). Another focus of studies on reasoning is linear syllogism, also referred to as the three-term series problem. When the structure of a mental model is the same as the structure of what it represents, the model is said to be iconic. Iconic mental models have the advantage of yielding conclusions that are not a part of the premises used in their construction (emergent consequences). Furthermore, people tend to construct the simplest mental model possible (parsimony), which can sometimes lead to reasoning errors. In contrast to the mental models theory, the natural deduction systems approach to reasoning argues that people follow deduction rules to draw conclusions from propositions.
In order to further study reasoning, Wason invented many reasoning tasks, including his generative problem, which refers to a problem where an individual must generate information in order to solve it. In such problems, individuals must incorporate a strategy to falsify hypotheses, in order to eliminate incorrect beliefs. This is known as an eliminative strategy. However, Wason observed that participants rarely do this, and instead are inclined to use a confirmation bias, meaning that they seek information that confirms—rather than disconfirms—their hypotheses. Although deceptively simple, Wason’s most significant contribution was his selection task, which involves deducing the validity of a rule from four cards. Wason’s experiments then raised a debate as to how people come to reason.
Much of the above research examined reasoning as a single logical system. Cosmides, on the other hand, argued that different types of problems require different types of “domain-specific” procedures. An example of domain-specific reasoning is the finding that individuals reason according to social contracts.
Reasoning can be viewed as drawing a valid conclusion based on given information. Making judgements and choices, however, requires us to choose among possible, sometimes uncertain, out-comes. Research on judgment and choice was pioneered by the heuristics and biases program of Tversky and Kahneman, which describes rules governing—and the systematic introduction of errors into—the reasoning process. It states, for instance, that people make assumptions about the occurrence of an event based on the law of large numbers. Similarly, people mistakenly rely on a law of averages, which can result in gambler’s fallacy. People can also make errors when relying on the law of small numbers because it is incorrect to assume that all small samples are representative of a population (representativeness heuristic). Errors can also occur when people become overly biased to an initial value when asked to judge the magnitude of something (anchoring and adjustment heuristic). Moreover, the availability bias influences individual judgment because people have a tendency to believe that more available instances occur more often. In addition, it has been argued that availability can lead to an illusory correlation. However, Shweder maintained that correlation is not an intuitive concept, and is illustrated by how people reason about events that are mathematically known as the regression to the mean.
In response to the heuristics and biases program, it was argued that, rather, people use simple heuristics that depend on ecologically valid cues, such as the recognition heuristic. Utilizing such a heuristic is ecologically rational. On the other hand, according to the less-is-more-effect, people can sometimes make better judgments if they know less. However, these ideas have also been criticized, with some authors suggesting that the recognition heuristic is limited in its ability to generalize too many situations, and that the simple heuristics are limited in their ability to be applied to truly important and difficult decisions.
Chapter Objectives
- To describe various understandings of syllogistic reasoning.
- To identify various heuristics and biases that affect people’s judgments.
- To understand ecological rationality.
- To explore the importance of training in statistical reasoning.