None
Thursday, 23 May 2013

Everybody's talking about MOOCs (Massive Online Open Courses) but what's it like to take one?

I decided to find out.

In March I completed an eight week course in Data Analysis taught by a bio-statistician at John Hopkins University. It used the Coursera Platform.

There were approximately 102,000 students enrolled in the course, about 51,000 watched videos, 20,000 did quizzes, and 5,500 did everything. ‘Everything' meant watching the videos, doing the quizzes and assignments and also grading the assignments using the peer review process.

I was one of the 5,500 and found that I spent between 8 and 20 hours per week on the course. I was on sabbatical which was just as well as the course had tight deadlines with weekly content and quizzes which we received on Sunday and had to be completed a week later.

What were the highlights of the course?

The discussion boards were great. With that many people on the course it is easy to get help and fun to give it. People went to extraordinary lengths to explain things even developing alternative examples and R code to help me understand when I struggled with an idea. All the help came from other participants.

The Teaching Assistants were mainly there to police the board and to make sure people weren't giving the answers out and to deal with technical issues like not being able to download data files. I found that when I didn't grasp how one person had explained something, that others would jump in and try alternative approaches. Sooner or later someone would have an explanation that I could follow.

The peer-peer grading was an interesting experience. We had to write a formal report (Abstract, Introduction, Method, Results, Conclusions and References) for both assignments. After submission we had to grade four of our peers using a marking key. Following this we had to grade our own report. I then had the opportunity to compare what I had scored myself against the average score from my four assessors. In the main, it was very close.

I did the course for a number of reasons. I needed to learn R, a statistical programming language and the course would force me to do so. I also wanted to develop some competence in data analysis so I could collaborate more effectively with my colleagues in the Statistics department. Finally I wanted to see what all the hype of MOOCs was about.

As a result of taking the course I'd also like to try some of the teaching processes they used. In particular: a) short recorded lectures followed by quiz questions that require you to use to material in the lecture to solve a specific problem; b) having four tries at answering the quiz question (I learned more when I made a mistake than when I got it right first time);and c) peer-peer marking. Of course all of these require an appropriate software platform which is one of the selling points of Coursera.

I'd also like to be able to see all the data that the course generated, see which quiz questions we struggled with, what people did when they struggled, all of this is captured by the software.

Though we would need data analysis skills to analyse it of course!

For those interested in seeing what the lectures are like, the course videos are on Youtube and tagged by week: https://bit.ly/16PPtuI ; and the course lecture notes are on Github: https://github.com/jtleek/dataanalysis

Tags

Groups
UWA Forward