SILC Showcase

Showcase November 2015: Integrating Cognitive Science with Innovative Teaching in STEM Disciplines: Spatial Learning in STEM

Share this article using our url:

Integrating Cognitive Science with Innovative Teaching in STEM Disciplines: Spatial Learning in STEM

Jennifer Cole, Matt Ford, David Miller, Kay Ramey, Jue Wu, Robert Linsenmeier

Northwestern University

In September, Northwestern University held a small conference about spatial skills in the undergraduate curriculum. Speakers represented cognitive science, natural sciences, and engineering aspects of spatial learning. Many STEM faculty in the audience were interested in the research findings as they apply to their students, but were not experts in this domain, which prompted many conversations. Speaker Jonathan Wai (Duke University) identified spatial reasoning as a domain equal in importance to math and verbal skills for success in STEM, but not often evaluated in standardized tests. Other speakers presented data on students’ capabilities in spatial learning, the impact of their learning on persistence in STEM fields, interventions that can improve spatial skills, and issues in spatial visualization.

Figure 1
Tim Shipley illustrates testing a skill that geologists need in order to see beneath the surface at Integrating Cognitive Science with Innovative Teaching in STEM Disciplines.

The conference was sponsored by Northwestern University through the McCormick School of Engineering and Applied Science and the Northwestern Center for Engineering Education Research, the Weinberg College of Arts and Sciences, and the Associate Provost for Undergraduate Education, as well as the Spatial Intelligence and Learning Center.

Nora S. Newcombe (P. I.), Laura H. Carnell Professor of Psychology at Temple University, opened the meeting with a discussion on the current state of research in spatial learning. Examples from discipline-based education researchers (DBERs) and cognitive scientists were compared, suggesting that the best approach to understanding students’ abilities in spatial reasoning lies in a combination of the two approaches, thereby incorporating proven instructional practices in the classroom while generating the useful STEM classroom-based evidence to determine which parts of a classroom intervention were actually successful. She and Thomas F. Shipley, Associate Professor of Psychology, Temple University, have tried to deepen their understanding of cognitive processes in spatial reasoning by working in detail in the domains of geology and geography. Results from a case study in which students read topographical maps were presented. Hand gestures, pointing and tracing, and language were used to direct students in interpreting the elevation lines of a topographical map. This is difficult for students, but when these approaches were paired there was an improvement in their spatial skills. Tim Shipley also discussed working to understand how professional geologists use spatial information in their work and how they avoid the erroneous conclusions that novices would make.

David H. Uttal, Professor of Psychology and Education at Northwestern University, addressed the role that spatial ability plays in predicting STEM achievement. Studies from multiple STEM disciplines suggest that spatial ability serves as a gatekeeper for students entering STEM, but matters more for problem-solving success among STEM novices than it does among experts, at least for basic disciplinary problems. Mike Stieff, Associate Professor of Learning Science and Chemistry at the University of Illinois at Chicago picked up this theme. It is often suggested that students’ spatial abilities are predictors of whether students can master understanding of 3D spatial topics such as orbitals, spectra, molecules, bonds, and rotations. However, a comparison of expert and novice chemists showed that experts use strategies other than the ones that students rely on early, such as mental rotation, and that moving students toward the strategies that experts use can help students be more efficient in solving problems.

While it may sometimes be possible to find other strategies to solve spatial problems, spatial skills are still important in many fields. Both a recent meta-analysis on spatial training by Uttal, and work by Sheryl Sorby, Professor of STEM Education, Ohio State University, with undergraduate engineering students, showed that spatial ability can be improved, and that this improvement is associated with positive STEM outcomes. Sorby discussed her experience in developing a spatial skills course for students performing poorly on the Purdue Spatial Visualization Test (PSVT). Her intervention, implemented over several years at Michigan Tech, has effectively improved students’ spatial skills, and thus led to better grades in engineering graphics, and also in other introductory STEM courses (e.g. chemistry, engineering) , including some where spatial skills might seem irrelevant (e.g., pre-calculus, computer science). Her intervention seemed to be particularly important for the women engineering students who were at risk of leaving engineering.

Figure 2
David Uttal, at Integrating Cognitive Science with Innovative Teaching in STEM Disciplines, arguing for research to move beyond correlations between spatial intelligence and STEM performance to understand why the correlation does, or in some cases, does not, exist.

Amy Shelton, Professor and Associate Dean for Research in the School of Education at Johns Hopkins University, shared findings from a program of research on developing spatial skills among both undergraduate and K-12 students. Supporting Sorby’s work, she showed evidence that an optional, hands-on spatial training course for freshman engineering students, using a variety of tangible and digital puzzles and games, led to improvements in performance for students with low spatial scores on the PSVT. She presented preliminary evidence comparing students taking that spatial course with students taking a different course that was intended as a control, which also had small enrollment and involved teamwork and more attention from the instructor, but whose content was not spatial practice. The course was also beneficial in improving confidence and persistence, and suggests that some of the effects (but not the increase in spatial ability) might be due to the social attributes of such a course.

Sketching and visual representations were discussed in several of the presentations. Kenneth D. Forbus, Walter P. Murphy Professor of Computer Science and Professor of Education at Northwestern University described CogSketch, a computational tool which parses natural sketches for deep meaning. CogSketch mimics human spatial reasoning to draw inferences and generate conceptions of function or meaning. For example, students were asked to illustrate various features and processes at a geological fault line, and CogSketch graded their submissions and gave natural language feedback based on its "understanding" of the student's intent. CogSketch has applications both as an educational tool, helping students reason through spatial problems, and as a research environment for understanding human cognition. Wade Goodridge, Assistant Professor of Engineering Education at Utah State University, explained how people with good spatial skills have a much easier time conceptualizing and designing almost any engineering project. He reflected on how his own experiences in sketching, drafting, and vocational arts prepared him to do things that some of his peers without those experiences found difficult, and to move on to engineering. Andrew Heckler, Associate Professor of Physics at Ohio State University, addressed the relationship between spatial reasoning and displays in the domain of physics, with a particular emphasis on vectors. In his study, he compared students’ abilities to use two different representations of spatial vectors – the traditional visual arrow representations and algebraic representations. Students’ abilities to add and subtract vectors were generally worse when using arrows and better when using algebra, suggesting that reasoning in the algebraic format is the more fundamental skill, while the arrow format is the more advanced one, which surprised many in the audience. Interestingly, when using the algebraic format for solving problems, students do just as well at visualizing the resultant vector as if they had started with a visual representation. Stephanie M. Gardner, Assistant Professor of Biological Sciences at Purdue University, spoke of the frustrations with getting students to generate appropriate graphical representations or interpret graphs in a biology course. Her team created a classroom intervention and rubric to measure graphical communication and graph choice. Following the principles of Tufte, her goal is to get students to create graphs that have substance, economy, complexity, and honesty, and to integrate the graphical representation with verbal and statistical ones for a richer understanding.

Sarah W. Bednarz and Robert Bednarz, Professors of Geography at Texas A&M University discussed frameworks for classifying types of spatial problems (how many different spatial skills are there?), and methods for going beyond the PSVT to assess students’ spatial skills. One study assessed spatial skills along five dimensions of spatial habits of mind: pattern recognition, spatial description, visualization, spatial concept use, and spatial tool use, while another study developed a taxonomy of spatial thinking based on three dimensions: concept, representation, and cognitive process.

You are here: SILC Home Page SILC Showcase Showcase November 2015: Integrating Cognitive Science with Innovative Teaching in STEM Disciplines: Spatial Learning in STEM