story by Claire Miller
College students’ experiences in introductory science, technology, engineering and math (STEM) courses can have a major impact on whether they decide to pursue a degree in STEM.
“Research shows that approximately half of the students who initially aim to earn STEM degrees do not achieve their goals,” said Min Kyu Kim, associate professor in the College of Education & Human Development’s Department of Learning Sciences. “Students from underrepresented groups often experience increased concern about their abilities in STEM fields, especially after receiving lower grades. Therefore, it is crucial to provide support for undergraduate students in STEM majors to help them achieve successful learning experiences during their early college years.”
Kim is the principal investigator on a three-year, $298,375 National Science Foundation grant to help students from underrepresented backgrounds gain STEM knowledge and prepare for future success in their classwork.
He’s working with M. Shammer Abdeen, co-principal investigator and physics lecturer in the College of Arts and Sciences, to integrate and test the Student Mental Model Analyzer for Research and Teaching (SMART) in two introductory physics classes at Georgia State University.
Students in these courses are asked to work on specific activities outside of class, such as completing reading assignments and watching pre-lecture videos, to prepare for upcoming class sessions.
SMART is a web-based, artificial intelligence-supported system that allows students to submit summaries of what they’ve learned in these pre-class activities, revise their work based on SMART’s feedback and take quizzes to test their new knowledge.
From here, the faculty members can log into SMART’s dashboard to see how individual students scored and adjust their upcoming in-class lesson plans based on students’ performance.
“For instance, if a pre-lecture video receives low SMART averages, indicating a lack of comprehension among students, the instructor can revise the next lecture to include more information and explanations,” Kim explained. “By identifying concepts that students struggle with, instructors can allocate more time to those areas and employ suitable grouping strategies, such as discussion, debate, hands-on labs or problem-solving.”
The grant project will give Kim and Abdeen data on how students progress through their pre-class assignments and help them identify the physics topics that students benefit most from having SMART’s feedback on. It will also help the researchers determine how pre-class activities and students’ work in the SMART system prepares them for tackling more demanding tasks during class.
As the grant continues, Kim and Abdeen will train other Georgia State faculty members, part-time instructors and graduate teaching assistants how to use SMART in the introductory physics courses. They also plan to share SMART and its learning modules with faculty members at other two-year and four-year colleges to expand the number of students who can benefit from this technology.
Kim hopes this work will offer more insight into how pre-class activities and the AI-based SMART system can support underrepresented students interested in STEM.
“Extensive data collection will uncover the relationships between pre-classroom activities, in-classroom performance, self-efficacy, interest in physics and student demographics – such as gender, race, ethnicity and first-generation status – over time,” Kim said. “Ultimately, this project will lay the groundwork for future research aimed at developing an AI-scaffolded pre-classroom learning model that promotes success for the majority of students in introductory physics courses.”