Logged In or Logged Out? A Statistical Analysis on Absences and Virtual Presence in Business Statistics

Abstract

The COVID-19 pandemic forced faculty and students to adapt to a culture of virtual learning. The transition back to normalcy was immediately positive for some, while others maintained preferences towards the virtual classroom. We elaborate on the importance of classroom attendance in business statistics courses with an online learning component by estimating the impact of virtual learning on performance. We use a sample of six sections of business statistics across three semesters. Each semester employed the same instructor, text, lecture format, and algorithmic assignments but differed in delivery. One semester was completely virtual, one was mixed in-person and virtual, and one was completely in-person. We find that students who attended class virtually scored significantly lower on exams than their classmates who attended lectures in person. Additionally, students who attended virtually and were absent for one or more classes performed worse on exams when compared to students attending in person. Online assignment performance is a strong predictor of exam success but less so for students who attend class virtually.