I was tempted to dismiss this paper (27 page PDF) out of hand when I read it was investigating "a new scenario: the fully online large university class." After all, we've been doing this for a few decades now. But I liked the oveall approach of the study involving, essentially, the study of student engagement rather than simple behavioural measures (like course grades) to evaluate course quality. Importantly, grades are not a measure of quality, and engagement is not a predictor of grades. "We suggest moving from the exclusive use of a prediction algorithm based only on behavioral data to the analysis of richer data sources that explore more types of engagement (behavioral but also emotional and cognitive)."
Today: 1 Total: 85 [Direct link] [Share]
