Cognitive load and representational competence: The development of an adaptive learning system to assist students with structure creation

Many science, technology, engineering, and mathematics (STEM) students are required to take chemistry courses, learn to construct representations to help them visualize the structures of molecules that make up matter, and use these representations to understand the properties of matter. This structure-creation process is often difficult for students and may be hindered by issues associated with 'cognitive load', which is the global amount of mental effort connected to a task. In instances where the overall cognitive load is high, students may struggle with completing the task and learning may be adversely impacted. The goals of this project are to study fundamental issues surrounding the role that cognitive load plays in undergraduate chemistry students' ability to construct representations, to use the results of these studies to develop an adaptive learning system that will help students become more proficient at creating chemical representations, and to assess the impact this system has on students' construction abilities and learning of chemistry. This work has the potential to enhance students' proficiency with chemical representations, and therefore, help them gain a more thorough understanding of key chemistry concepts that promote success in STEM.

This NSF Improving Undergraduate STEM Education (IUSE: EHR) project seeks to better understand the interplay between cognitive load and students' abilities to construct representations of chemical structure. It constitutes one of the first formal research studies to use physiological metrics such as heart rate and electroencephalography to measure cognitive load as undergraduate chemistry students work to solve problems. Unlike other methods of measuring load, these techniques can be gathered in real time and have been shown, in research conducted in other disciplines, to be responsive to small changes in load. Results of these studies will then be used to design and develop an adaptive learning system that will focus on helping chemistry students construct Lewis and skeletal structures of molecules. The system will make use of an initial diagnostic system that is informed by the research results to ascertain the students' construction ability and provide Socratic-style feedback as they seek help in creating their representations. Although the system developed for this project will be specifically designed for the chemistry classroom, the general methodological approaches developed will likely have broad applicability to other STEM disciplines and serve as models for similar work in those fields.