Qing X Ryan
Associate Professor, Physics & Astronomy, College of Science
Research
What is Physics Education Research?
Physics Education Research (PER) is an interdisciplinary field with a wide range of research areas that can involve cognitive science, psychology, education, neuroscience, etc.; and uses a variety of research methodologies that includes both qualitative and quantitative methods. Here I list some examples of my research areas, students are also encouraged to explore their own interest and come up with their own project ideas.
Social psychological variables and students' performance:
Social psychological measures such as students’ interest, motivation, utility value towards physics could affect students’ learning outcomes. We explore whether the relationship between membership in an under-represented demographic group and course grades are mediated by psychological measures such as Interest, Math anxiety, etc.
Student Difficulites in Upper-division Physics
Student difficulties at both the introductory and upper-division level have been broadly investigated by the PER community. Upper division problem solving involves more complicated math and more sophisticated physics topics. A growing body of research suggests that upper-division students continue to struggle with problem-solving in these advanced physics topics. This project investigates student difficulties in upper-level physics courses, with a focus on topics appear several times in different contexts across the advanced undergraduate physics curriculum. We recently explored student's use of boundary conditions in the context of electromagnetism and quantum mechanics. We are interested in investigating students' use of diagrams during problem-solving.
Some previous projects:
Using machine learning to predict students’ performance in introductory courses
Other than the statistical method used in the previous study, machine learning is another type of data analysis method. Both of which are used in “Educational data Mining” (EDM) to draw conclusions from large educational datasets. With machine learning and data science as a whole are growing explosively in many segments of the economy, we now have more tools to make sense and exploit the exponentially growing data (especially those collected in an increasing online world). In this project, we use random forest machine learning algorithms to predict students’ performance in introductory physics course using many background variables including pre-requisite course gpa, in-class assignments performance such as homework scores, and demographic variables.
Research-based Assessments for Upper-division Physics
Research-based conceptual assessments play an important role in physics education research. Research-validated instruments are used to characterize common and persistent student difficulties, as well as to support curricular transformation. This project is trying to develop such an assessment for upper-division electrodynamics course, using an objectively-gradable format to improve scability.
C3PO: Customizable Computer Coaches for Physics Online
Solving complex, non-routine problems in context has been recognized as essential for all citizens of a modern society, but nowhere are they more important than for scientists and engineers. In collaboration with the University of Minnesota, the C3PO project is working towards helping physics students develop a generalizable, expert-like, problem-solving framework by creating a set of online coaching programs that supports problem solving.