A randomized controlled experiment for comparing face-to-face and online teaching during COVID-19 pandemic

Randomized controlled experiments have shown that face-to-face teaching is more effective in delivering various learning outcomes than asynchronous online teaching. Unlike the asynchronous online teaching mode, the synchronous online mode has a live instruction component and is more comparable to the face-to-face mode. A small-sized randomized controlled experiment involving 50 students showed that there was no significant difference in student ratings on the effectiveness between the face-to-face and synchronous online teaching modes. Prior to the current study, no medium-or large-sized randomized controlled experiment had been conducted for comparing the two modes. The current study aims to fill in the gap by comparing the effectiveness of face-to-face (i.e., intervention) and synchronous online (i.e., control) teaching through a randomized controlled experiment involving 725 students from seven statistics courses offered by the Department of Statistics and Actuarial Science at the University of Hong Kong. Results show that the difference in learning outcomes between the two modes is not statistically significant. The class size is an effect modifier that students assigned to the face-to-face mode have significantly higher final weighted and final exam scores if they have face-to-face lessons with 25 students or fewer. The Pass/Fail grading option has a significantly negative effect on course performance.