A while back, we did an analysis of the number of participants that was, it turns out, based on flawed data. Unknown at the time was that a back-end change in how the e-mails were being addressed after the grouping process removed them from the campaign data that mailgun uses to generate the charts. This means unique opens and unique clicks are almost certainly higher than what the graph indicates. How much higher is hard to say.
All things considered, it’s probably as straightforward just to look at the number of opens for each of the assignment e-mails so far (the ones at the beginning of each week that contain the actual work to be done):
- Week 1: 6,926
- Week 2: 3,747
- Week 3: 1,418
- Week 4: 2,180
- Week 5 (so far): 1,175
It’s likely a safe assumption that some learners are opening the e-mails more than once, but these e-mails again would seem to point to an engaged learner population above 500 or so, and one that seems to be fairly stable after week 3. But we’ll keep poking around for additional data points.