Developmental Differences in Young Children’s Solutions of Logical vs. Empirical Problems
Bradley J. Morris,V. Sloutsky
DOI: https://doi.org/10.4324/9781410603494-81
2020-12-22
Abstract:We examined the development of the ability to differentiate logically determinate from logically indeterminate problems. The results indicated that a) young children tend to reduce the number of empirical possibilities via “cutting” the second half of less informative propositions, b) these errors do not stem from encoding or recall errors, c) from elementary to middle school, children tend to increase their understanding of logical form, and d) this increase corresponds to a decrease in the rate of cuts. There is a large body of research examining children’s understanding of empirical indeterminacy (Fay & Klahr, 1996; Piéraut-Le Bonniec, 1980; Sodian, Zaitchik, & Carey, 1991). A problem is empirically determinate if it corresponds to exactly one empirical possibility; otherwise, it is empirically indeterminate (Piéraut-Le Bonniec, 1980). Previous findings suggest that young children often (1) fail to appreciate empirical indeterminacy, confusing indeterminate problems with determinate ones, but not vice versa, and (2) have less difficulty solving determinate problems than indeterminate problems (Bindra, Clarke, & Schultz, 1980; Byrnes & Overton, 1986; Fay & Klahr, 1996; Pieraut-Le Bonniec, 1980). However, there is another class of problems that should be considered in conjunction with the issue of determinacyproblems that require logical, but not empirical solutions. These problems are logically determinate (LD) if they are solved logically, but they are indeterminate if they are solved empirically. Researchers have demonstrated that children (and many adolescents) do not fully understand logical determinacy (Byrnes & Overton, 1986; Moshman & Franks, 1986; A. Morris & Sloutsky, 1998) and they often attempt to provide empirical solutions to logically determinate problems (A. Morris & Sloutsky, 1998; Osherson & Markman, 1975). In this article, we examine the development of solution strategies to some logical and empirical problems and possible cognitive mechanisms underlying these strategies. Information-processing analysis of solving logical vs. empirical problems Logically determinate problems are those that can be solved a priori based on their logical form. Some LD problems yield logically true or necessary conclusions, whereas others yield logically false or impossible conclusions. Problems that yield conclusions that are true in some, but not all states of affairs are defined as logically indeterminate (LI), or empirical. It has been traditional since the early work of Newell & Simon (1972) to conceptually model problem solving as search through a problem space for a desired goal state. A three-stage model outlines the creation of problem space, search and creating an output. Encoding is the creation of problem space from the information in the environment. The more clearly the problem space represents salient elements in the environment, the more veridical the representation, and the greater opportunity the organism has for solving the problem (Newell & Simon, 1972; Newell, 1990). The second phase is search in which a decision matrix is examined for possible outcomes and actions of the represented problem. Two types of search are utilized: problem search, in which search proceeds through possible outcomes of states and operators, and knowledge search, in which search is through memory (Newell, 1990). Once a goal state, an impasse, or some terminating point in search is reached, an output is then created. Therefore, it seems safe to assume that the mentioned reasoning and problem solving errors may occur due to the following factors (or any of their combinations): 1) Limits on encoding. 2) Poor representation of problem space. 3) Incomplete search through problem or memory space. 4) Inaccurate mapping of a problem solution onto a verbal response. Evidence from several related domains such as scientific reasoning, logical reasoning, and practical reasoning indicates that children and adults often limit their search in both problem space and memory space (Kuhn et al., 1995; Markovits, 1988; Mynatt, Dohetry, & Tweney, 1977; Tweney et al, 1980). They were also found to exhibit a ‘positive capture’ strategy failing to consider equally plausible alternatives (Bindra et al., 1980; Fay & Klahr, 1996). The tendency to limit their search has also been found in studies of logical reasoning in that both children and adults tend to not search for counterexamples in forms such as conditional reasoning (Johnson-Laird, 1993; Markovits, 1988; Wason & Johnson-Laird, 1972). In practical reasoning (when the task was to compare size of foreign cities) people often made fast and frugal “Take-the-Best” inferences relying on a small number of the most salient cues (such as the familiarity of the city), while ignoring the rest of the cues (Gigerenzer & Goldstein, 1996). There is also evidence that young children exhibit difficulties when solving problems corresponding to no empirical possibilities, such as contradictions (B. Morris, 1998; Sloutsky, Rader, & Morris, 1998). In particular, when presented with a contradiction, preschoolers responded as if they ignored (or “cut”) the second half of the contradiction thus transforming it into a statement corresponding to one empirical possibility functioning to limit problem space. Therefore, it seems plausible that young children implicitly assume that propositions correspond to exactly one empirical possibility, thus creating a “defective” or incomplete problem space. If this is the case, then the number of empirical possibilities compatible with the problem could predict the problem difficulty. The easiest problems are those that correspond to exactly one empirical possibility whereas an increase or decrease in the number of empirical possibilities leads to an increase in problem difficulty and subsequently to the number of errors (for a detailed discussion see Sloutsky, Morris, & Rader, in review). Three groups of children (preschool, elementary and middle school) were presented with reasoning problems. The experiment focused on (a) the ability of children and adolescents to distinguish logical from empirical problems; (b) solution strategies for different types of problems; (c) patterns of errors; (d) the relationship between the number of empirical possibilities and the problem difficulty; and (e) accuracy of encoding and mapping of verbal responses. We deemed it necessary to reduce the number of possible sources of error via eliminating the necessity of search through knowledge space. In so doing, we presented participants with knowledge-lean problems that required “deriving a solution from givens” rather than requiring memory search. Method Participants 38 four-and fiveyear-old children enrolled in three child care centers (average age = 4.3 years; 16 girls and 22 boys), 34 third grade children in three elementary school classrooms (average age =8.4 years; 19 girls and 15 boys), and 35 sixth grade children enrolled in two middle school classrooms (average age = 11.7 years, 16 girls and 19 boys).