Traditional tests of concept knowledge generate scores to assess how well a learner understands a concept. Here, we investigated whether patterns of brain activity collected during a concept knowledge task could be used to compute a neural ‘score’ to complement traditional scores of an individual’s conceptual understanding. Using a novel data-driven multivariate neuroimaging approach—informational network analysis—we successfully derived a neural score from patterns of activity across the brain that predicted individual differences in multiple concept knowledge tasks in the physics and engineering domain. These tasks include an fMRI paradigm, as well as two other previously validated concept inventories. The informational network score outperformed alternative neural scores computed using data-driven neuroimaging methods, including multivariate representational similarity analysis. This technique could be applied to quantify concept knowledge in a wide range of domains, including classroom-based education research, machine learning, and other areas of cognitive science.

Traditional tests of concept knowledge generate scores to assess how well a learner understands a concept. Here, we investigated whether patterns of brain activity collected during a concept knowledge task could be used to compute a neural 'score' to complement traditional scores of an individual’s conceptual understanding. Using a novel data-driven multivariate neuroimaging approach—informational network analysis—we successfully derived a neural score from patterns of activity across the brain that predicted individual differences in multiple concept knowledge tasks in the physics and engineering domain. These tasks include an fMRI paradigm, as well as two other previously validated concept inventories. The informational network score outperformed alternative neural scores computed using data-driven neuroimaging methods, including multivariate representational similarity analysis. This technique could be applied to quantify concept knowledge in a wide range of domains, including classroom-based education research, machine learning, and other areas of cognitive science.

is a Lecturer at Dartmouth's Geisel School of Medicine where she teaches biostatistics and research methods and designs educational assessments, courses, and curricula. One of her research areas is what constitutes competence in engineering work. Based on her research, she designed a performance review system for Ph.D. level engineers regarding their technical and professional competencies. She has also taught statics and dynamics to engineering students.

No abstract

Distributed neural systems engage in coordinated information processing that develops over time to support complex learning. Here, we extract the information represented in these neural systems in order to observe changes in conceptual knowledge related to STEM learning. Two groups of learners with different levels of prior knowledge and experience completed an fMRI-based task involving abstract mechanical engineering concepts. First, using data collected during learning, we identified emergent networks of neural activity that displayed similar response patterns. Next, we extracted the representational structure of these informational networks using multivariate representational similarity analysis. Finally, by comparing these representations to an expert model of concept knowledge, we identified within each informational network the presence (or absence) of expert-level knowledge of the target concepts. Results demonstrate that between groups and over the course of learning, different neural systems represent expert concept knowledge, indicating distinctions between dorsal and ventral stream processes and along an anterior-posterior gradient within the ventral stream. Our approach can be applied to investigate conceptual change across STEM domains, and provides a novel means of assessing the neural basis of concept learning over time and between individual learners.