Coursera again – customization and computation

Hearing Daphne Koller speak about Coursera this evening at the University of Pennsylvania really brought into high relief for me the ideas of Clay Christensen about using technology to do two seemingly contrary things:

1) customize learning for individual learners and,
2)  scale it — as a customized experience — to huge numbers of people.

Skillfully using “computational methods” to customize teaching and learning is key for higher education, for libraries.

Because survival, in my opinion, will not come from exclusion, selectivity, making people queue up for your services.  It will come from finding a business model that allows you to be hugely inclusive, and vastly lower the barrier (the “cost” to people) to benefit from your services.

One way to lower “cost” is to have very good customization.  If I am helping someone learn something, the less of their time I waste the better.  Time is wasted whenever teaching and learning is outside the zone of the individual student’s proximal development — it’s too easy or too hard.  Getting a grip on that zone has been sort of an art.  In the library world, we call it the reference interview, and it’s a conversation we often have with users to figure out where they are cognitively, what they do and don’t know about research and research tools.  Traditionally, we’ve done that one-on-one.  Not really scalable (even on a campus with 20,000 students, let alone three times that many.)

Since I became actively involved in learning management systems a few years back, I have been increasingly interested in using online teaching as a way to help students develop information skills.  A lot of what I’ve seen librarians do in blended or online learning is weak – we tend to put links to useful resources into course web sites, or to transfer our one-on-one or group session practices from F2F venues into the online world.  We haven’t used technological affordances to transform our teaching – we’re doing the same things we always have (essentially handouts with links and Q & A) but in a different venue.

Offering googobs of links is not pedagogy.  I would compare this to the early days of MIT’s courseware, where “stuff” was made available, but it wasn’t instructional unless you wanted to put a whole lot of individual effort into it.  A list of potentially useful library resources in a course web site is not, obviously, instructional, and it’s a pretty low level of customization (at the course level rather than the student level).  We can do better, but it matters a lot how we frame the task.

One of the keys for computational solutions is to carefully define the problem you need to solve. An ” art” or something mystical and magical that seems to happen when a librarian helps a student develop research skills doesn’t lend itself to a computational solution.  Computational solutions exist where we can specifically define the types of skills needed, and the knowledge gaps that exist, and design our learning experiences around these articulated problems.  Over time, using learning analytics will enable us to better target our customizations for individuals. But our first task is to work with faculty and students to carefully analyze how students using research tools differ in their typical approaches from what we would consider expert behavior.

We’re not starting from zero – some of this analysis has already been done.  Almost all my conversations with faculty who assign library research, for example, mention the “five perfect articles” syndrome.  Students early in their college careers often begin research by looking for articles that perfectly match their own research question.   They find it challenging to gather relevant information by understanding how each research question has different facets, and how comprehensively understanding the scholarly conversation on your topic involves looking at the literature in these different facets.  This is directly connected to the gap between the way students do keyword searching in a database and the way experts do it.  Expert keyword searching is driven by understanding this facets issue.  Sadly, very many librarian-led search workshops don’t look at the facets problem at all – what they teach is to “think of synonyms and try them.”   That “Just try another one” approach is not helping students learn research as a form of critical thinking.  Imagine if you bought clothes that way — you’d just wander through the clothing department trying on every single pair of pants without ever thinking about why some styles are more flattering than others!

We can do better.  We understand the “five perfect articles” problem and its relationship to the relevant aspects of a research question – now what we need to do is design an engaging and useful online learning environment and put it in the path of students so that they can’t get their research done without experiencing the learning sequence we have designed.

And then go on to the next specifically defined skill/knowledge gap…

–Karrie Peterson


About Karrie Peterson

Head, Liaison, Instruction & Reference Services
This entry was posted in Disruptive Technologies, Learning, Learning Analytics, Shifting Education Paradigms and tagged , , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s