The Developmental Psychology of Reasoning and Decision-Making

Henry Markovits

• Publisher: Psychology Press
• Year: 2013
• eBook ISBN: 9781317931065
• Topic: Psicología
• Subtopic: Psicología cognitiva y cognición

Logical thinking is a critically important cognitive skill. It is not just essential for mathematical and scientific understanding, it is also of prime importance when trying to navigate our complex and increasingly sophisticated world. Written by world class researchers in the field, The Developmental Psychology of Reasoning and Decision-Making describes the ways that children learn to reason, and how reasoning can be used to overcome the influence of beliefs and intuitions.

The chapters in this edited collection focus on the new, revolutionary paradigm in reasoning and cover the recent research on the development of reasoning in two important areas:

• Cognitive abilities required to reason well and how these abilities develop in children and adolescents.
• Recent empirical data showing the effect intuition and prior belief have on reasoning, even when the outcome is inappropriate.

Different theoretical and empirical perspectives from recent Piagetian theory, mental models and gist processing are examined, along with empirical results looking at specific aspects of reasoning in children. The key theme of the book is to better understand how reasoning develops not only through examining 'logical' reasoning, but also the nature of the interactions between people's intuitions and their reasoning abilities.

The Developmental Psychology of Reasoning and Decision-Making provides an overview of the main theories and key empirical results related to the development of reasoning and should be of particular interest to students and researchers in developmental psychology and education, along with those in cognitive psychology.

1Introduction

Henry Markovits

Reasoning and decision-making are at the core of advanced thinking. Determining how these develop over time is thus critical for both educational purposes and for our theoretical understanding. In the following chapters, we have a state-of-the-art overview of the latest developmental approaches. Although there is no current consensus as to the exact nature of the underlying processes used for advanced thinking, there has been a striking change in the way that these processes have been articulated that illuminates the complexity of human reasoning. Initial developmental studies of reasoning were motivated by Piaget’s theory. The stage of formal reasoning was characterized by the ability to reason logically, and was considered to be the end-point of development. In this perspective, development could be considered to be a long but consistent march towards better and more logical thinking. This in turn was considered to involve the gradual independence of reasoning from “empirical knowledge and beliefs”, leading to the ability to use essentially abstract rules and structures.

Many, although not all studies during this period did find increases in the ability to reason logically with age. Developmental research thus often contrasted the Piagetian position with the idea that younger children were more “logical” than anticipated and that advanced thinking was really accessible at a much earlier age. These studies were done with the underlying assumption that adults were indeed logical thinkers, just as Piagetian theory would predict. At the same time, studies of adult reasoning and decision-making were showing that even well-educated people have great difficulty in making logical decisions. From the influential heuristic and biases program of Kahneman and Tversky to reasoning studies deriving from Wason’s seminal observation that even doctoral students are not able to analyze a seemingly simple selection task in a logical way, it has been convincingly shown that people very often reason in what appears to be completely illogical ways. Thus, even adults who should be in the formal stage of reasoning are strongly influenced by empirical knowledge and belief when reasoning.

The tension between these two sets of results is clear; they highlight the necessity of understanding the interactions between processes that can lead to logical thinking and those that lead to use of heuristic, belief-based reasoning. Dual-process theories have attempted to codify these effects by distinguishing between System 1 intuitive, rapid forms of reasoning and System 2, abstract rule-governed processes. The most straightforward reading of this approach would suggest that children should be more prone to intuitive reasoning leading to logical mistakes and that, with age, the impact of System 2 processes should increase. However, as pointed out in several chapters in this book, this simply does not correspond to what can be observed; the developmental pattern of interactions between heuristic factors and logic is in fact quite complex.

Given the strong tendency of adults to make intuitive judgments that are not logically, or rationally appropriate, it is clear that a simple replacement model whereby irrational, intuitive processes are gradually superseded by more complex forms of logical reasoning must be wrong. This raises some interesting questions which in one way or another underlie much of current developmental approaches to reasoning. The first of these is the epistemological status of intuition. There is a natural tendency to presume that any kind of rapid heuristic reasoning must be inherently illogical. However, there are sensible biological grounds to suppose that intuition, while possibly more variable than more explicit forms of analytic processing, must be reasonably rational in some sense. If this is the case, then the trade-off between heuristic and analytic processing must be more complicated at any age, since both forms of reasoning can be considered to have both benefits and costs. Thus, it becomes possible to argue, as we see in Chapter 3 by Weldon, Corbin and Reyna and Chapter 7 by De Neys, that intuitive processes allow access to some form of logical reasoning. In contrast, logic and rationality can also be conceived as the domain of explicit higher-level forms of processing, as witnessed by Moshman’s metacognitive analysis or Gauffroy and Barrouillet’s mental model approach.

A second, related question has to do with the degree of independence of intuitive and analytic processes. Once again, a simple version of the intuitive versus analytic dichotomy would claim that the two systems do not overlap in the way that they use information. This would suggest that intuitive processing relies on surface characteristics and heuristic shortcuts, while analytic processing requires a deeper semantic analysis. However, the sheer variability of logical reasoning and the fact that some research suggests that younger children are not susceptible to the same degree of heuristic biases as adults suggests that both processes might share some common sources. The interaction between emotion (which can be seen as a form of heuristic) and reasoning is clearly seen in Beck and Riggs’ analysis of counterfactual reasoning.

Thus, it has become increasingly clear that understanding the nature and development of reasoning and decision-making requires looking at the specific processes that can explain how both familiar and abstract forms of knowledge can be derived from what might be a common base, which includes both learned information and biologically based tendencies, and importantly how these processes can interact in a complex developmental progression. This has necessarily led away from the logical or not dichotomy that has often characterized developmental studies towards more process-oriented analyses that have the potential to explain how the interactions between empirical knowledge, belief and logic develop. The following chapters represent an invaluable synthesis of the different ways that these interactions can be currently understood.

These chapters can be divided into two categories. Part 1 provides large scale syntheses and theory. Chapter 2 by Toplak, West and Stanovich distinguishes between intelligence and rationality. These authors make the critical point that reasoning and judgment require understanding how these relate to some criteria of rationality, which involves using intelligence in a much more complex way than is usually imagined. They provide an analysis of rationality based on the fluid versus crystalized dichotomy and give an exhaustive and original survey of the judgment and decision-making field that makes an excellent starting point for any analysis of the developmental literature. Their approach suggests that while rationality cannot be equated to intelligence, the two do co-vary in some basic sense. Weldon, Corbin and Reyna in Chapter 3 provide a very different analysis of the interaction between intelligence and rationality. Gist theory claims that intuition underlies rationality; more specifically, that intuitive processes allow extraction of logical rules from complex information leading to what can be called an intuitive logic. In this perspective, conscious processing of information often leads to misleading analysis of surface characteristics, missing the underlying rationality. An opposing theory is provided in Chapter 4 by Gauffroy and Barrouillet’s adaptation of mental model theory to development. While allowing a clear role for heuristic processes, this theory considers that the chief component of development is the increase in working-memory capacity. Reasoning is seen as a conscious and effortful attempt to process the underlying semantics of propositions by the construction of explicit models. Finally, Klaczynski and Felmban in Chapter 5 provide a detailed analysis of developmental patterns of heuristic responding, with emphasis on the role of developmental inversions. While not tied to a specific theory, this analysis provides a useful methodological overview of just what heuristic responding might mean, and how to reasonably measure it.

Part 2 provides more detailed analyses of the specific cognitive functions that must underlie development. Moshman in Chapter 6 highlights the importance of understanding both the general progression from specific inferential processes, which often remain unconscious, to a more reflective metacognitive form of reasoning that can explicitly consider truth as an explicit construct. However, he also points out the domain specificity of such reasoning, implying that while there might be some underlying general processes involved, it is impossible to understand how more complex forms of reasoning develop without understanding the specific constraints of each domain. De Neys in Chapter 7 reports the results of a recent, innovative research program that shows that people have an intuitive form of logic detection. More specifically, when making heuristic judgments that are not “logical”, both children and adults who are on the surface very comfortable with these responses, show unconscious signs of discomfort. Such a process is, of course, an important mechanism that could potentially provide a proximal mechanism by which intuition might lead to logical reasoning, even in very young children. This idea suggests that logic must be somewhat primitive, if it can be intuitively processed. In contrast, in Chapter 8 Markovits suggests that reasoning can be seen as motivated by the coordination of divergent thinking across multiple levels of abstraction. Reasoning is not considered as logic per se, but involves the ability to conceive of possibilities, with necessity requiring elimination of possibility. This model places emphasis not on logic as a rule-driven process, but as an increasingly abstract way of analyzing real-world knowledge. Finally, in Chapter 9, Beck and Riggs give a synthesis of their research program on the development of counterfactual reasoning. Such reasoning is not only a logical milestone, but represents a link with social and emotional worlds that clearly show the complexity of the relations between rationality and these other perspectives. They convincingly demonstrate that a dichotomous model does not correspond well to what is known about counterfactual reasoning, and that a more nuanced model that can account for a variety of different forms of competencies is required.

What these approaches share is an increasingly nuanced perspective about the way that different forms of information processing, rational and intuitive, concrete and abstract, interact in order to produce an adult who is sometimes capable of amazing feats of formal logic, and who at the same time makes judgments and reasons in curiously illogical ways. How children develop into such adults is critical for our understanding of what reasoning is and how to facilitate its acquisition. This book provides a synthesis of current understanding of the processes underlying this development.

Part I

Overview and theory

2Assessing the development of rationality

Maggie E. Toplak, Richard F. West and Keith E. Stanovich

Variation in intelligence has been one of the most studied topics in psychology for many decades (Deary, 2001; Geary, 2005; Lubinski, 2004), and the development of the cognitive abilities related to intelligence is likewise a central topic in developmental science (Anderson, 2005; Bjorklund, 2004). Because assessments of intelligence (and similar tests of cognitive ability) are taken to be the sine qua non of good thinking, it might be thought that such measures would serve as proxies for the developmental trajectories for judgment and decision-making skills. It is important to understand why such an assumption would be misplaced.

Judgment and decision making are more properly regarded as components of rational thought, and it is often not recognized that rationality and intelligence (as traditionally defined) are two different things conceptually and empirically. Distinguishing between rationality and intelligence helps explain how adolescents can be, at the same time, intelligent and irrational (Reyna & Farley, 2006; Stanovich, 2006). Thus, the developmental trajectories of the cognitive skills that underlie intelligence and those that underlie rational thinking must both be studied in their own right because they are conceptually and empirically separable, as we will argue in the next section. Judgment and decision-making skills, as critical components of rational thought, have a developmental trajectory that cannot just be inferred from the development of general cognitive ability.

Distinguishing rationality and intelligence in modern cognitive science

Cognitive scientists recognize two types of rationality: instrumental and epistemic. The simplest definition of instrumental rationality is: behaving in the world so that you get exactly what you most want, given the resources (physical and mental) available to you. Somewhat more technically, we could characterize instrumental rationality as the optimization of the individual’s goal fulfillment. Economists and cognitive scientists have refined the notion of optimization of goal fulfillment into the technical notion of expected utility. The model of rational judgment used by decision scientists is one in which a person chooses options based on which option has the largest expected utility (Baron, 2008; Dawes, 1998).

The other aspect of rationality studied by cognitive scientists is termed epistemic rationality. This aspect of rationality concerns how well beliefs map onto the actual structure of the world. Epistemic rationality is sometimes called theoretical rationality or evidential rationality (Manktelow, 2004; Over, 2004). Instrumental and epistemic rationality are related. The aspects of beliefs that enter into instrumental calculations (that is, tacit calculations) are the probabilities of states of affairs in the world.

One of the fundamental advances in the history of modern decision science was the demonstration that if people’s preferences follow certain patterns (the so-called axioms of choice—things like transitivity and freedom from certain kinds of context effects), they are behaving as if they are maximizing utility. They are acting to get what they most want (Luce & Raiffa, 1957; Savage, 1954). This is what makes people’s degrees of rationality measurable by the experimental methods of cognitive science. Although it is difficult to assess utility directly, it is much easier to assess whether one of the axioms of rational choice is being violated. This has been the logic of the seminal heuristics and biases research program inaugurated in the much-cited studies of Kahneman and Tversky (1973, 1979; Tversky & Kahneman, 1974; see Kahneman, 2011).

Researchers in the heuristics and biases tradition have demonstrated in a host of empirical studies that people violate many of the strictures of rationality and that the magnitude of these violations can be measured experimentally. For example, people display confirmation bias, they test hypotheses inefficiently, they display preference inconsistencies, they do not properly calibrate degrees of belief, they overproject their own opinions onto others, they combine probabilities incoherently, and they allow prior knowledge to become implicated in deductive reasoning (for summaries of the large literature, see Baron, 2008; Stanovich, 2009, 2011). These are caused by many well-known cognitive biases: base-rate neglect, framing effects, representativeness biases, anchoring biases, availability bias, outcome bias, vividness effects, and various types of attribute substitution (Kahneman & Frederick, 2002), to name just a few. Degrees of rationality can be assessed in terms of the number and severity of such cognitive biases that individuals display. The important point, however, is that none of these processes are assessed directly on intelligence tests.

Conceptually as well, intelligence (as actually measured) concerns cognitive components quite different from the judgment and decision-making components that define human rationality. Intelligence, as measured on many commonly used tests, is often separated into fluid and crystallized components, deriving from the Cattell/Horn/Carroll (CHC) theory of intelligence (Carroll, 1993; Horn & Cattell, 1967). Sometimes termed the theory of fluid and crystallized intelligence (symbolized Gf/Gc theory), this theory posits that tests of mental ability tap, in addition to a general factor (g), a small number of broad factors, of which two are dominant (Geary, 2005; Horn & Noll, 1997). Fluid intelligence (Gf) reflects reasoning abilities operating across of variety of domains—in particular, novel ones. It is measured by tasks of abstract reasoning such as figural analogies, Raven Matrices, and series completion. Crystallized intelligence (Gc) reflects declarative knowledge acquired from acculturated learning experiences. It is measured by vocabulary tasks, verbal comprehension, and general knowledge measures. Ackerman (1996) discusses how the two dominant factors in the CHC theory reflect a long history of considering two aspects of intelligence: intelligence-as-process (Gf) and intelligence-as-knowledge (Gc).

In our t...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• List of contributors
• 1. Introduction
• Part I: Overview and theory
• Part II: Development of specific competencies
• Index