Forecasting Exam Scores Based on How Students Participate in Introductory Geoscience Courses
Monday
3:00pm
REC Center Medium Ice Overlook Room
Oral Presentation Part of
Geoscience Education Research I
Authors
Perry Samson, University of Michigan-Ann Arbor
Ben van der Pluijm, Henry Ford Community College
This presentation reports on research being conducted at the University of Michigan on 1) how student participation is related to student grades and 2) how students participate as a function of their incoming GPA. This paper reports on the use of the Echo360 Active Learning Platform that resulted in data about student participation before, during and after class. Student participation measurements included per class information on attendance, volume of notes taken, number of instructor question answered, number of gradable questions answered correctly, number of questions asked and minutes viewing lecture captures. These data were coupled with the students' academic history from the university's student information system, survey results and exam grades to create a database with which to explore relationships between student participation and student outcomes.
Results show that 1) some strong relationships exist between student participation and student outcomes and 2) the nature of student participation is related to the students' incoming Grade Point Average (GPA). The quantity and quality of student responses to questions posed in class were found to be strong predictors of student grades on exams. These results offer a basis for constructing predictive models for student performance that can be implemented early in the academic term. Access to data at this level of granularity is considered an important step towards offering very personalized feedback to students based on their patterns of class participation.
Results also show that students with different GPAs entering the course participated in markedly different ways. Lower GPA students tended to participate in questions less often, took five times fewer notes and were more likely to participate in class remotely rather than physically come to class. These results suggest that student outcomes may have less to do with students' innate cognitive capability and more to do with poor motivation and/or study habits.
Results show that 1) some strong relationships exist between student participation and student outcomes and 2) the nature of student participation is related to the students' incoming Grade Point Average (GPA). The quantity and quality of student responses to questions posed in class were found to be strong predictors of student grades on exams. These results offer a basis for constructing predictive models for student performance that can be implemented early in the academic term. Access to data at this level of granularity is considered an important step towards offering very personalized feedback to students based on their patterns of class participation.
Results also show that students with different GPAs entering the course participated in markedly different ways. Lower GPA students tended to participate in questions less often, took five times fewer notes and were more likely to participate in class remotely rather than physically come to class. These results suggest that student outcomes may have less to do with students' innate cognitive capability and more to do with poor motivation and/or study habits.