My Metadata MOOC is over! I clicked the button to issue Statements of Accomplishment and Verified Certificates earlier today. (If you were a student in the MOOC and you’re reading this, it will take about a week for these to actually be issued, during which Coursera will complete their review and work their magic on their side.) I’m very pleased with how it went, but I’ll make that the topic of another post. Right now, I want to present some statistics about the course.

Coursera provides an embarrassment of riches to the instructor, where data about the course is concerned. We have stats about student activity, stats about views & downloads of the videos, stats about in-video quizzes and homeworks submitted, stats about participation on the discussion forums. We collected some, but by no means all, of this data, while the course was ongoing. And we posted the data we collected to the discussion forums, because we thought that the students might like to see the State of the Course in more or less real-time. (Where “real-time” is a word which here means “weekly.” Note: weekly video content was always launched on Monday mornings, and we collected & posted this data on the following Friday.) That was well received on the forums, and I’ll definitely be doing that again, for the second offering of the course. I recommend it to all MOOC instructors. Taking an online course can feel isolating. The discussion forums were very active, which certainly combatted the feeling of isolation. But I figured, any mechanism to help convey that this course is ongoing right now, live, as we speak, would help. And I’ll be honest, I really enjoyed watching these stats over time. But then, I’m a geek.

Anyway, this is the first of what will probably be several posts, where I report on the Metadata MOOC. First, let’s look at student activity:

I know that hard data on MOOCs is fairly thin on the ground. So for all you data wonks out there, here’s the raw data:

[table “” not found /]

The Coursera help documentation defines these categories as follows:

- Total Registered Students: “The number of (unique) students currently registered for this session, excluding those who have unregistered.” Some wag compared registering for a MOOC to clicking the Like button on Facebook. Total Registered Students is, to me, not a very meaningful statistic, since there’s no penalty for an individual to register for a MOOC and then never participate. If this stat means anything at all, it’s as a measure of interest.
- Total Active Students: “The number of (unique) students that have logged in at least once to the session site.” This is, for my money, a far more meaningful statistic: the number of individuals who actually did
*anything*in the course. - Active Students Last Week: “The number of (unique) students who logged on to the session site in the last week.” So that’s a rolling week, counting backwards from whenever the stats were generated.

It is interesting, but not surprising, that Total Registered Students decreases over the span of the course. I hypothesize that these un-registrants fall into 2 categories:

- Individuals who registered and then didn’t participate in the course, who got annoyed that they kept getting auto-generated emails about the course when we posted announcements, and so un-registered to shut those up.
- Individuals who participated in the course up to a point, realized that the course wasn’t as interesting to them as they thought it would be, or life intervened and they had to stop participating, or whatever, and dropped out.

It’s also interesting, but not surprising, that Total Active Students plateaued by around week 3. This makes sense when you consider that the homeworks in the course had deadlines: they were due about 2 weeks after the content for that unit launched. So if you started participating in the course during week 3, you’d already missed the deadline for the homework for unit 1. It would therefore be impossible for you to earn a Statement of Accomplishment, which required completion of all graded assignments.

This is something that I need to reconsider for the next offering of the course: Whether homeworks should have deadlines, or should be available until the end of the course. The former model forces students to form a huge cohort, proceeding through the course more or less in parallel. That seemed to work well this time around, and, I’d hypothesize, kept the discussion forums active. The latter model would allow students to do the course more at their own pace, and could enable more students to complete the course. I got a few annoyed emails & comments in the discussion forums about homework deadlines. But by and large, students seemed to be ok with the deadlines. I’m tempted to go with the latter model on the second round, just as an experiment. But that’s not a pedagogically sound reason. Anyone have thoughts on this, one way or the other?

The stat Active Last Week / Total Active is of my own invention. What percentage of all students in the course have participated recently? It’s a simple, but I think reasonably effective, measure of current activity in the course. I have to admit, I’m pretty pleased that I ended with about a quarter of all students still actively participating. I’ve had class sessions in the classroom that weren’t that active.

Stay tuned, folks, over the next few days, for more reporting about the Metadata MOOC. I’m just trying to keep these posts short enough that you’ll actually read them, plus give myself and the team at UNC some time to do some analysis.

## 13 Comments

## Henry Mensch

One of the blessings of a MOOC is that you can do it at your own pace … if you choose to do this then you don’t benefit from the cohort that’s pressing ahead at a quicker pace, but you benefit from what they’ve left behind (discussions, etc.). I don’t see a reason not to offer both options: select one on your way in, understand the implications of the option you’ve chosen, and then get started. Until MOOCs start awarding real credentials (college credit?) I’m not sure that the pace matters much.

I didn’t think the discussions were all that great, but I think that has more to do with the really diverse crowd which was attracted to the course.

## Willem van Gemert

I find the deadlines for the homework useful as reminders. During the week I was not able to follow the course in parallel with my work, so I followed the course and did the homework in the weekends. Without the deadlines I would probably not have completed the course. I found the homework (too) easy. I was always able to get a 100% score after 2-3 iterations. You have 10 in total … A colleague of mine accomplished the homework each time without actually having seen the videos which is a real shame and not the aim of the MOOC …

## Henry Mensch

I did exactly as you did … well, I also did something else: I ran through the homework once at first (and I only did this near the end) to determine what I already knew and what I needed to watch for … then I watched the videos and did the embedded quizzes (which mostly worked for me–occasionally there’d be a question which I couldn’t sort out) … then I did the homework “for real.”

## Scott Petri

Willem, have you considered that many course participants may come to the class with a high level of subject matter knowledge and may not need to access the lecture videos? They may just be surveying the course to see if their knowledge is still cutting edge, or in sync with the market. Others may be having their first exposure to the subject during this MOOC and may benefit from being able to practice and submit the homework multiple times. It is hard to be all things to all people, but that is what these courses strive to do. Like most aspects of education, the lesson that is hardest for students to learn is that they can only get back what they are willing to put in. Respectfully, Scott

## Data about the Metadata MOOC, Part 2: Video Viewership | Jeffrey Pomerantz

[…] ← Data about the Metadata MOOC, Part 1: Student Activity […]

## Data about the Metadata MOOC, Part 4: Completion | Jeffrey Pomerantz

[…] about for “dropout rate” is 90% (or more) of students who register for a MOOC. But as I’ve written, Total Registered Students is not a very meaningful statistic, since there’s no penalty for an […]

## An Answer to the Quetion of MOOC Attrition Rates - Degree of Freedom

[…] recently summarized some of his findings in a series of four blog entries which start here. And while even the author admits that his work so far represents a starting vs. an end point for […]

## #MOOC Who are my students? « Petridishing

[…] Pomerantz has blogged extensively about his Medadata students. In total, 27,623 students were active after […]

## A tale of two MOOCs | Querying libraries

[…] a tiny minority of 5%. To take another measure, of 25867 registered students in Week 8, only about half of them (14130) were “active” (defined as having logged into the MOOC), and only 3334 had been active in the final week of the […]

## Predictions about the Metadata MOOC round 2 | Jeffrey Pomerantz

[…] But that’s 8 weeks away. So to keep myself entertained in the meantime, I thought I could make some predictions about how student engagement in the course will play out over the next 8 weeks. My null hypothesis is that student engagement in the second round of the course will play out the same as the first round. By which I mean, I predict that the percentage of active students in the course will decline at the same rate in round 2 as it did in round 1. […]

## The MOOC Metrics We Should Be Tracking - Degreed Blog

[…] “Data about the Metadata MOOC, Part 1: Student Activity,” Jeffrey Pomerantz, July 10, 2014 “MOOCs @ Edinburgh 2013 – Report #1,” Edingburgh Research Archive, July 10, 2014 “HarvardX Working Papers,” Harvard University, July 10, 2014 “MITx Working Papers,” MIT Office of Digital Learning, July 14, 2014 […]

## What Do We Know About MOOC Students So Far?: A Look At Recent User Data | Free Education

[…] MOOCs provided via Coursera. Professors such as Jeffrey Pomerantz, who teaches a MOOC on Metadata, has blogged about his student activity. Providers, though sensitive about their numbers around student attrition, don’t seem to prevent […]

## atozhajjandumrah.com

There’s been a lot of talk about some of the dramatic numbers on MOOCs, but which ones are actually representative?