Specialization

Developmental Psychology

Courses Taught

• PSY 101 - Introductory Psychology
  
• PSY 301 - Child Development
  

Current Research

Deductive reasoning underpins everyday cognitive processes such as categorization as well as higher-order skills such as mathematical and scientific reasoning. My research focuses on cognitive development, specifically the role of learning in the development of deductive reasoning. Deductive reasoning is the process of drawing inferences that are necessarily true and includes evaluations of verbal descriptions, judgments about when evidence is needed and when it is not needed, and syllogistic inferences. In particular, I am interested in how children represent logical statements, how they use evidence and how they induce rules about the role of evidence for solving deductive statements. The development of logical reasoning has typically been studied as either a cognitive or language issue and nearly always as the result of innate or early-acquired processes that do not require learning. However, because logic is nearly always presented linguistically, an understanding of the mechanisms of cognition and language is required for an accurate account of logical performance. Thus, my research has focused on two broad areas: (1) logic and language and (2) mechanisms of logical development such as strategy acquisition and change.

Logic and Language

The development of logical reasoning is inextricably tied to the development of language because connectives such as AND, OR, & NOT have both a logical and conversational meaning. How do children come to know what such connectives mean? The fundamental issue is how natural language becomes formalized into an abstract reasoning system.

1. How do children make use of semantic information provided with negations? I examined whether preschool children's understanding of the word NOT is derived from their understanding of semantic information such as the concept of opposites. Children were asked to create pictures (or evaluate pictures created by other children) of simple negations, some using words with well-known opposites (e.g., up/down) and others without such an opposite (e.g., purple). For example, children were given a sticker in the shape of an arrow and the statement "The arrow is not pointing up." Children typically created a single picture that corresponded to the opposite however, when asked to evaluate pictures, children agreed that non-opposites (e.g., arrow point right) were true, suggesting they understood negations beyond simple opposites. Thus, although children can represent more than one possible match, they had a preference for a one-to-one correspondence between representation and semantic information. We call this a "minimalist representation" in which children do not represent all available information but represent only enough information to produce a putative solution.

2. What is the role of experience in children's developing understanding of logical connectives? Little is known about the prevalence of logical connectives in young children's language environment. I have coded over 100,000 natural language interactions between parents and children to determine the relative frequency of the connectives AND, OR, & NOT to determine the relationship between connective use and the development of logical reasoning. The coding system examines the speaker, frequency, and the pragmatic or didactic use of connectives. (Preliminary) analyses indicate that children hear NOT approximately twice as frequently as AND, and 25 times more frequently than OR. The relative frequencies suggest significant differences in the amount of experience children have with each connective. The data also indicate that parents adjust the proportion and pragmatic use of connectives to approximate the level at which their child uses a particular connective, a process known as finetuning.

Mechanisms- Strategy Change

We know that children's logical reasoning performance improves across development within three areas of abilities: (a) to represent logical statements, (b) to evaluate evidence, and (c) to determine when evidence is necessary. But how do children develop knowledge of each area? And how are these areas integrated to form an understanding of logic? This line of research examines how connectives become formalized into logical reasoning. Statements with these connectives differ in how evidence is evaluated and when evidence is necessary. Statements such as tautologies (always true) and contradictions (always false) do not require evidence and are logically determinate. All other statement types are logically indeterminate. These statements require evidence but differ in the number of true and false "states" with which they are associated. The following questions guide my program of research:

1. How do children represent logical statements? In a series of studies, I examined the relationship between the type of statement and the number of possible 'states' children represent for each. For example, I presented children with a set of pictures of six objects and asked them to match the picture to a description such as "The triangle is blue AND the triangle is not blue." Children could either match the description to a specific object (e.g., a blue triangle), a question mark that indicates they believe there is more than one match, or a trash can that indicates they believe there is no match. Children preferred matches to a specific object even when such a match did not match the truth-value. The results support the idea that, as on the connective level, children prefer "minimalist representations" on the statement level.

2. How do children use evidence to solve deductive problems? Correctly evaluating evidence requires two processes: (1) correctly mapping the evidence, and (2) assigning the correct truth-value (true, false, can't tell). I found that children often made a systematic error when evaluating evidence. For example, when asked to evaluate a statement like "The ball is red and the ball is not red" children often said "true" when shown a red ball and "false" when shown a green ball. This error pattern occurs when children disregard the connective and the second part of the statement. We have labeled this common error pattern a "cut". I recently examined other types of possible error patterns in evidence evaluation. In this experiment, I gave children a series of statements and offered them the opportunity to request evidence to evaluate these statements. I have proposed a set of response patterns that corresponded to evaluation strategies.

3. How do children acquire rules about the role of evidence for solving deductive statements? Complete processing of logical statements requires correct evidence evaluation and knowledge of when evidence is necessary or unnecessary. In another study, I presented children with a series of determinate and indeterminate statements and asked each child (up to) three probe questions about each statement: (1) to evaluate the statement before evidence, (2) if they would like to request evidence, and if requested, (3) to evaluate the evidence. I proposed a series of strategies and a set of possible response patterns derived from these strategies. For example, someone with a full understanding of logical reasoning should evaluate a contradiction as "false" before evidence and refuse evidence, because it is unnecessary for evaluating the statement. Many children requested evidence for all statements and made frequent errors evaluating evidence. Overall, the results suggested that the order of processing acquisition is nearly opposite the order of processing; children are able to correctly evaluate evidence before they correctly request or refuse evidence. Age-related changes in processing phases suggest that experience with logical statements results in the elimination of redundant processing steps (strategy compilation) and the creation of new knowledge categories (elaboration).

To investigate the compilation-elaboration hypothesis, I am currently running a training study in which third and fifth grade children are given a brief instruction in logical reasoning. In the mapping condition, children are given training that directed attention to both parts of a statement and its logical connective. Children in the necessity condition are given the training provided in the mapping condition and are given a series of probe questions about whether evidence is necessary in each case. Preliminary results indicate that the children in the necessity group dramatically increase performance in both evidence evaluation and knowing when evidence is necessary, whereas children in the mapping condition increase performance in evaluation only. The results are consistent with the compilation-elaboration hypothesis.

Summary and Goals

The goal is an empirically grounded theoretical model of logical development that incorporates children's experience with language, social interaction and education. The model itself should be computationally plausible, detailing mechanisms of knowledge acquisition and change (e.g., compilation and elaboration) that is grounded in the natural language experience of children. In particular, I am interested in further examining the role of experience in the development of logical reasoning. For example, one project, will examine the relationship between how children of different ages process a statement based on the statement's believability. A second line of research examines individual differences in strategy use. A third line of research is the development of a computational model of strategy change in ACT-R in which the proposed mechanisms of change can be instantiated and empirically evaluated. A related line of research will further investigate the role of language experience in logical reasoning. Specifically, this line of research will examine the influence of input on children's performance on reasoning tasks and further examine the nature of the social context in which children hear and produce logical statements. I am also interested in the application of this research-based model to educational settings. Once I have found procedures that have been effective in 'laboratory' settings, I would like to begin translating these procedures for use in a 'classroom-based' setting by working with teachers to develop effective classroom-based procedures for teaching basic logical reasoning by itself and as a foundation for science and math skills.

Representative Publications

Morris, B. J. (In press). Opposites attract: The influence of predicate dimensionality on preschool children's interpretations of negations. Journal of Child Language.

Morris, B. J., & Schunn, C. D. (in press). Rethinking the development of logical reasoning skills from a strategy perspective. In R. Maxwell & E. Newton (Eds.), Methods of thought: Individual differences in reasoning strategies. Psychology Press.

Morris, B. J., & Sloutsky, V. M. (2001). Children's solutions of logical vs. empirical problems: What's missing and what develops? Cognitive Development, 16(4), 907-928.