5.4 Using Student Behavior Data from a Survey Course to Design Interventions for Low-Achieving Students

Tuesday, 9 January 2018: 2:15 PM
Ballroom C (ACC) (Austin, Texas)
Perry J. Samson, Univ. of Michigan, Ann Arbor, MI

Variations in “educationally purposeful” in-class digital student behaviors have been compared with student exam grades in a large, blended atmospheric sciences survey course to determine which, if any, behaviors were best related to exam grades. The study, which included six semesters over three years (N=1,347) represented a blended course wherein students could participate synchronously either in the classroom or remotely with technology that invited and expected student participation. These data were collected in the context of incoming grade point average (GPA) to examine both how student digital behaviors were related to exam grades and how students with different incoming grade point averages behaved differently in the course.

Results show that student behaviors were significantly correlated with exam scores. Notably, the correctness of their answers on formative assessments were related to grade as might be expected, but also the percentage of student participation in formative assessment activities and measures of attentiveness including slides viewed during class were also statistically related to exam grades. The strongest relationship with grades was found to be the students’ incoming grade point average but incoming grade point average was in turn shown to be strongly related to student behaviors in the course. That is, students who had performed poorly in the past tended to carry poor participation behaviors into this course.

On the other hand, it was found that students with lower incoming GPAs who chose to participate at a higher level in this course received higher grades. Moreover, this work was able to quantify the gains in student exam scores as a function of increasing digital participation within various incoming grade point average cohorts. This work serves as a basis for designing personalized evidence-based interventions based on prior success and student behaviors.

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner