Saturday, October 29, 2011

Open Educational Video Content Viewing

I taught an intermediate microeconomics course last spring for which I developed (and am continuing to develop) online content.  All of the content, with the exception of assessments delivered in Moodle quizzes, can be found on the Economics Metaphor blog.  Much  of that content are videos done in short modules.  These are screen captures of Excel simulations with my voice over and then that is captioned.  The YouTube channel for those videos shows snips of each video with title and link,  a modest number of subscribers, and the the number of views per video.  Since I don't typically track this stuff I don't know if any past subscribers have since unsubscribed, or if subscribers come to the content they watch via their subscription or other ways.  In any event, the info shown on the channel page is for the entire history and therefore doesn't tell you that much.

I'm not sure what possessed me to do the following, but this morning I started to look at the stats YouTube generates for content creators, where there is some control of time interval for which data is being reported.  A creator can look at such data on a video by video basis or for all videos together.  Below I will report about data from the second week in August till yesterday, for all videos.   My students from the spring have little to no reason for watching the videos during this period.  So I'm pretty confident that this represents other viewing. 

There were not quite 21,000 views in total since I started posting and about 7,500 views in the two and a half months that we're focusing on.  None of the videos have gone viral.  But there is enough information to draw some conclusions.  There are 53 videos in the channel.  Of those, 5 are not economics content related and thus unlikely to generate traffic.  (A couple give instructions on how to navigate the Excel simulations.  One is an example of voodoo in the technology - the Jing screen capture distorted the image terribly.)  So there are 48 videos that might have generated the views.

Here are some screen shots of the reports YouTube generates.  This first one shows very low traffic prior to my course last spring, then a drop off in the summer after the course had concluded, then an upswing this fall with a usage rate seeming higher than in the spring.  It also shows how the viewers are finding the videos - mainly via search.  I found it interesting that the general Google search didn't generate that much traffic.  The bulk of the views come from within YouTube, either via YouTube search or from related videos. 

This second one shows the geographic distribution of the viewers.  About 45% come from the U.S. and then no other country has even 10%.  Though the captions in the videos can be translated,  my guess is that very few of the viewers can't understand English, based on this geographical pattern.  The second also shows the top 10 traffic producers from among the videos in the channel.   The top 5 videos generate a bit more than 50% of the traffic.  From this I conclude that the viewers are mainly students.  They are looking for videos on specific topics.  They are not looking to this content as a full course offering.  These videos must be complementing or substituting for specific content from the intermediate micro course they are taking, perhaps because they find the topic particularly hard to understand. 

I'm not sure what to make of this third one, which shows the age distribution and gender of those viewing the content.   Where does that info come from?  The profiles in YouTube of these viewers?  If so, we should ask whether people enter correct profile information.  So, with a skeptical eye, I will say these results are quite surprising.  The viewers are overwhelmingly male and the bulk are middle aged.  There are comparatively few who are traditional students as measured by age.  If these data are accurate, then I have to reconsider my conclusion that most viewers are students.  This info is more consistent with the viewers being economics instructors. But I would have thought that instructor wouldn't want to see individual modules but rather a cluster of modules that build to an understanding of a topic, which is how videos were put together for the spring class. 

Let me wrap up.  We are talking about video content that people are actively looking for.  Most don't stumble on this completely out of the blue.   Some "tests" of how much they value what they find potentially could be available from looking at how their subsequent search behavior for video content is impacted by having found a particular video.  A very rough test is whether they make subsequent searches at all at later dates.  In other words, having done this in this past, are they more likely to do a similar sort of looking in the future?  A more refined test is what search terms are used in the subsequent searches.  Do they start to search by creator or creator concatenated with a specific topic?  Since related videos (many by other creators) provide a significant referral function, the way YouTube offers up those related videos in the right sidebar must impact which videos get watched.  

Is this type of demand for content something that should engage creators or others on the campuses where the creators work?   I don't know, but possibly the answer is a strong affirmative.  With online learning, so much of the thinking has been about entire courses and quite frequently with those the content resides in an LMS or campus hosted content server, which may entirely preclude this sort of external discovery or, if not that, then not actively encourage it.  Particularly for those campuses that have a strong outreach mission, perhaps there is a big opportunity loss in taking that approach. 

No comments: