home - login - register

PERC 2018 Abstract Detail Page

Previous Page  |  New Search  |  Browse All

Abstract Title: Towards a model of students pathways in STEM
Abstract: Understanding what factors affect changes to students' majors, what leads to students leaving STEM, or what indicates that a student will leave the university helps us better define the boundary conditions students operate within and leads to data driven approaches to undergraduate advising. We have built a machine learning model using transcript data that predicts who stays in physics and who switches to an engineering program (Accuracy=92.3%). This model demonstrates students who switch frequently take engineering courses prior to switching, and frequently do not take the third semester course in modern physics. Performance in introductory physics/calculus courses, gender, and ethnicity all play smaller roles in the model. This may imply that students who leave for engineering planned to do so when they entered physics, that students enter physics programs and are attracted away by engineering courses, or that students taking initial physics courses may become disenfranchised as physics students (amongst other reasons). These results set clear research questions that can be investigated using qualitative research methods.
Abstract Type: Contributed Poster Presentation
Session Time: Poster Session II
Poster Number: B2

Author/Organizer Information

Primary Contact: John M. Aiken
University of Oslo, Center for Computing in Science Education, Department of Physics
Sem Sælands Vei 24
Uio/Fysisk Inistitutt
Blindern, Non U.S. 0316
Phone: +16786973181
and Co-Presenter(s)
Marcos D. Caballero
Department of Physics
Michigan State University
1310A Biomedical and Physical Sciences Building
567 Wilson Rd.
East Lansing, Michigan 48824-1046
(770) 827-3185