Bryan J. Matlen

Neural Correlates of Fluid Reasoning in Children and Adults

Fluid reasoning, or the capacity to think logically and solve novel problems, is central to the development of human cognition, but little is known about the underlying neural changes. During the acquisition of event-related fMRI data, children aged 6–13 (N = 16) and young adults (N = 17) performed a task in which they were asked to identify semantic relationships between drawings of common objects. On semantic problems, participants indicated which of five objects was most closely semantically related to a cued object. On analogy problems, participants solved a visual propositional analogy (e.g., shoe is to foot as glove is to…?) by indicating which of four objects would complete the problem; these problems required integration of two semantic relations, or relational integration. Our prior research on analogical reasoning in adults implicated left anterior ventrolateral prefrontal cortex (VLPFC) in the controlled retrieval of individual semantic relationships, and rostrolateral prefrontal cortex (RLPFC) in relational integration. In this study, age-related changes in the recruitment of VLPFC, temporal cortex, and other cortical regions were observed during the retrieval of individual semantic relations. In contrast, age-related changes in RLPFC function were observed during relational integration. Children aged 6–13 engage RLPFC too late in the analogy trials to influence their behavioral responses, suggesting that important changes in RLPFC function take place during adolescence.

Memory suppression is an active process that improves over childhood

We all have memories that we prefer not to think about. The ability to suppress retrieval of unwanted memories has been documented in behavioral and neuroimaging research using the Think/No-Think (TNT) paradigm with adults. Attempts to stop memory retrieval are associated with increased activation of lateral prefrontal cortex (PFC) and concomitant reduced activation in medial temporal lobe (MTL) structures. However, the extent to which children have the ability to actively suppress their memories is unknown. This study investigated memory suppression in middle childhood using the TNT paradigm. Forty children aged 8–12 and 30 young adults were instructed either to remember (Think) or suppress (No-Think) the memory of the second word of previously studied word-pairs, when presented with the first member as a reminder. They then performed two different cued recall tasks, testing their memory for the second word in each pair after the TNT phase using the same first studied word within the pair as a cue (intra-list cue) and also an independent cue (extra-list cue). Children exhibited age-related improvements in memory suppression from age 8 to 12 in both memory tests, against a backdrop of overall improvements in declarative memory over this age range. These findings suggest that memory suppression is an active process that develops during late childhood, likely due to an age-related refinement in the ability to engage PFC to down-regulate activity in areas involved in episodic retrieval.

Semantic similarity of labels and inductive generalization: Taking a second look

Prior research suggests that preschoolers can generalize object properties based on category information conveyed by semantically-similar labels. However, previous research did not control for co-occurrence probability of labels in natural speech. The current studies re-assessed childrens generalization with semantically-similar labels. Experiment 1 indicated that adults made category-based inferences regardless of co-occurrence probability; however, 4-year-olds generalized with semantically-similar labels that co-occurred in child-directed speech (e.g., bunny-rabbit) but not with non-co-occurring labels (e.g., crocodile-alligator). Experiment 2 indicated that generalization with semantically-similar labels increased gradually between 4- and 6-years of age. These results are discussed in relation to theories of early learning.

Development of inductive generalization with familiar categories

Inductive generalization is ubiquitous in human cognition. In the developmental literature, two different theoretical accounts of this important process have been proposed: a naïve theory account and a similarity-based account. However, a number of recent findings cannot be explained within the existing theoretical accounts. We describe a revised version of the similarity-based account of inductive generalization with familiar categories. We tested the novel predictions of this account in two reported studies with 4-year-old children (N = 57). The reported studies include the first short-term longitudinal investigation of the development of children’s induction with familiar categories, and it is the first study to explore the role of individual differences in semantic organization, general intelligence, working memory, and inhibition in children’s induction.

Development of Category-Based Induction and Semantic Knowledge

Category-based induction is a hallmark of mature cognition; however, little is known about its origins. This study evaluated the hypothesis that category-based induction is related to semantic development. Computational studies suggest that early on there is little differentiation among concepts, but learning and development lead to increased differentiation based on taxonomic relatedness. This study reports findings from a new task aimed to (a) examine this putative increase in semantic differentiation and (b) test whether individual differences in semantic differentiation are related to category-based induction in 4- to 7-year-old children (N = 85). The results provide the first empirical evidence of an age-related increase in differentiation of representations of animal concepts and suggest that category-based induction is related to increased semantic differentiation.

Development of Category-Based Reasoning in 4- to 7-Year-Old Children: The Influence of Label Co-Occurrence and Kinship Knowledge.

Category-based reasoning is central to mature cognition; however, the developmental course of this ability remains contested. One strong indicator of category-based reasoning is the propensity to make inferences based on semantically similar labels. Recent evidence indicates that in preschool-age children the effects of semantically similar labels are limited to a small subset of labels that co-occur in child-directed speech, suggesting that performance with these labels may reflect lexical priming rather than category-based reasoning. However, most co-occurring labels used in prior research refer to offspring-parent relationships (e.g., puppy-dog). Thus, it is possible that children in previous research performed induction by relying on kinship rather than co-occurrence information. To address this possibility, the current studies examined the role of kinship knowledge and label co-occurrence in induction in 4- to 7-year-old children and adults. The results point to a gradual age-related increase in the ability to spontaneously rely on kinship knowledge when making inferences.

Promoting Sketching in Introductory Geoscience Courses: CogSketch Geoscience Worksheets

Research from cognitive science and geoscience education has shown that sketching can improve spatial thinking skills and facilitate solving spatially complex problems. Yet sketching is rarely implemented in introductory geosciences courses, due to time needed to grade sketches and lack of materials that incorporate cognitive science research. Here, we report a design-centered, collaborative effort, between geoscientists, cognitive scientists, and artificial intelligence (AI) researchers, to characterize spatial learning challenges in geoscience and to design sketch activities that use a sketch-understanding program, CogSketch. We developed 26 CogSketch worksheets that use cognitive science-based principles to scaffold problem solving of spatially complex geoscience problems and report observations of an implementation in an introductory geoscience course where students used CogSketch or human-graded paper worksheets. Overall, this research highlights the principles of interdisciplinary design between cognitive scientists, geoscientists, and AI researchers that can inform the collaborative design process for others aiming to develop effective educational materials.

Executive function in learning mathematics by comparison: incorporating everyday classrooms into the science of learning

ABSTRACT Individual differences in executive functions (EFs) are well established to be related to mathematics achievement, yet the mechanisms by which this occurs are not well understood. Comparing representations (problems, solutions, concepts) is central to mathematical thinking, and relational reasoning is known to rely upon EF resources. The current manuscript explored whether individual differences in EF predicted learning from a conceptually demanding mathematics lesson requiring relational reasoning. Analyses revealed that variations in EF predicted learning when measured at a delay. Thus, EF capacity may impact students’ overall mathematics achievement by constraining their resources available to learn from cognitively demanding reasoning opportunities in lessons. To assess the ecological validity of this interpretation, we report follow-up interviews with mathematics teachers who raised similar concerns that cognitively demanding activities such as comparing multiple representations in mathematics may differentially benefit their high versus struggling learners. Broader implications for ensuring that all students have access to, and benefit from, conceptually rich mathematics lessons are discussed. We also highlight the utility of integrating methods in science of learning (SL) research.

Mechanisms of Spatial Learning: Teaching Children Geometric Categories

Children’s representations of geometric categories like triangles are often centered on a prototypical exemplar (e.g., an equilateral triangle). New cases are judged based on perceptual similarity to the prototype; such a strategy leads to systematic errors in categorization. Creating correct geometric categories requires children to move beyond a reliance on perceptual similarity and learn category-defining rules (e.g., a triangle is an enclosed, three-sided shape). In this research, we test whether a brief training experience using comparison could help three- and four-year-old children learn the category of triangle. Further, we ask whether different types of comparisons (within-category or between-category) support learning in distinct ways. The data indicate that both types of comparison fostered category learning, but that within-category comparisons promoted generalization to new exemplars whereas between-category comparisons reduced overgeneralization to non-exemplars. Furthermore, these effects were moderated by the perceptual similarity of the compared pairs. The results indicate that comparison can foster spatial category learning.

Impact and Prevalence of Diagrammatic Supports in Mathematics Classrooms

Mathematical problem solving typically involves manipulating visual symbols (e.g., equations), and prior research suggests that those symbols serve as diagrammatic representations (e.g., Landy and Goldstone 2010). The present work examines the ways that instructional design of student engagement with these diagrammatic representations may impact student learning. We report on two studies. The first describes systematic cross-cultural differences in the ways that teachers use mathematical representations as diagrammatic supports during middle school mathematics lessons, finding that teachers in two higher achieving regions, Hong Kong, and Japan, more frequently provided multiple layers of support for engaging with these diagrams (e.g. making them visible for a longer period, using linking gestures, and drawing on familiarity in those representations), than teachers in the U.S., a lower achieving region. In Study 2, we experimentally manipulated the amount of diagrammatic support for visually presented problems in a video-based fifth-grade lesson on proportional reasoning to determine whether these multiple layers of support impact learning. Results suggest that learning was optimized when supports were used in combination. Taken together, these studies suggest that providing visual, temporal, and familiarity cues as supports for learning from a diagrammatic representation is likely to improve mathematics learning, but that administering these supports non-systematically is likely to be overall less effective.