Collaborative research: Neural and cognitive strengthening of conceptual knowledge and reasoning in classroom-based spatial education

Fostering true conceptual understanding in the minds of students, not just the right answers on the test, is the goal of teachers in every STEM classroom. Recent machine-learning advances in brain imaging allow detection of the neural signature of understanding STEM concepts. Critically, this lab-based technique has not yet been applied to learning in a real-world classroom. This method has transformative potential to complement traditional testing to identify education practices that yield the strongest concept learning. This method can also detect learned reasoning strategies. Given mounting evidence that spatial thinking is a driver of STEM success, this project will collect MRI and behavioral data before and after a high school geoscience course that uses a novel spatially-based curriculum to teach spatial concepts and spatial reasoning, and will measure learned changes in concept knowledge and spatial reasoning. Using an intervention that focuses on spatial relations (analogies), the project will test for the first time whether modifying a curriculum strengthens the representation of a concept in the brain and changes a neural approach to reasoning. Neural and traditional test measures will be combined to determine whether neural data help predict long-term strength of concepts and reasoning. Importantly, the vision is NOT that all students will one day be brain-scanned; it is to ultimately enhance the development of curricula by adding neural data to conventional testing in focus samples already used by curriculum developers. The major goals are to: 1) test whether real-world development of a concept and reasoning strategy can be measured at the neural level (as recently demonstrated in lab settings), 2) test whether neural markers have unique predictive value for long-term mastery (when combined with paper-and-pencil exams), 3) test neural measurement as a basis for guiding changes to a course curriculum, implementing a first test-case intervention that encourages students to form analogies between spatial relations, 4) develop a mechanistic theory of change for spatially-based education, including factors that impact disparity (e.g., sex and science anxiety). This proposal directly addresses the ECR research strands of Fundamental research on STEM learning, Research on STEM learning environments, and Research on gender in science, and has additional impacts for Broadening participation in STEM. Intellectual Merit. 1) Our cross-disciplinary team of experts in STEM classroom education, spatial cognition, reasoning, and neural bases of learning is committed to bridging the conspicuous gap between the neuroscience lab and the real-world classroom. This is not easy, but a confluence of advances in neuroimaging, and our partnership with Virginia school systems make this effort timely and tractable. To our knowledge, our ongoing NSF-funded project is the first to longitudinally measure how learning in a real-world high school changes the brain. Promising pilot data have emerged. What is needed next is the focus of the proposed project: to test the kind of neural representations (concepts and reasoning strategies) that can be used to actually inform instructional practice. Test performance is an indirect measure of underlying concept knowledge, and it is not always clear why scores go up or down. This project is motivated by the potential to complement traditional tests by directly measuring the strength of deep-lying concept representations. 2) Spatial thinking is a driver of STEM success. Our work with the “Geospatial Semester“