July 20th, 2017
posted by [syndicated profile] opencontent_feed at 05:28pm on 20/07/2017

Posted by david

I started to post this as a comment on Mike’s amazing essay Information Underload, but I’m going to put it here instead. Read Mike’s whole piece – it’s worth it.

He writes:

Endless thinkpieces have been written about the Netflix matching algorithm [including in education], but for many years that algorithm could only match you with the equivalent of the films in the Walmart bargain bin, because Netflix had a matching algorithm but nothing worth watching. (emphasis in original)

Is this why OER repositories (and the learning object repositories that came before them) typically fail – because the resource you find is frequently no better than the resource you could have made yourself if you had just spent the time creating instead of searching?

This contains echoes of the reusability paradox if you don’t understand that open licenses resolve the paradox. I suppose you could think about it from an information foraging perspective as well. But there’s some basic math around how we use time in relation to OER. If the time we spend searching for OER only turns up resources we could have created in roughly the same period of time, then there’s no advantage to OER. Being clear about that single point is super valuable. But Mike’s key insight here is that we shouldn’t try to solve this problem by decreasing mean time to discovery – we should solve it by increasing the value of the OER you eventually find.

Perhaps we should call this “OER leverage” – the ratio of time spent searching for OER to the time saved by finding OER.

As Mike says, “let’s belabor the point”:

  • Spending 15 minutes searching only to find an OER you could have created in about 15 minutes = not very useful
  • Spending 15 minutes searching and finding an OER that would have taken you 100 hours to create = very useful

This kind of example makes it clear that working to decrease mean time to discovery is a fight of diminishing returns. If I’m going to mostly find resources I could have made in 15 – 30 minutes, how much time can I possibly save by decreasing mean time to discovery? (Answer: 15 – 30 minutes.) There’s an upper bound on the amount of leverage I can achieve by working this side of the problem, and it’s a pretty low one. But if I work the other side of the problem – creating larger, more useful OER – there’s an opportunity to create significant leverage. How much time do I save when I discover a comprehensive set of OER that I can use to replace an entire textbook?

Mike continues,

Since Netflix is a business and needs to survive, they decided not to pour the majority of their money into newer algorithms to better match people with the version of Big Momma’s House they would hate the least. Instead, they poured their money into making and obtaining things people actually wanted to watch, and as a result Netflix is actually useful now…. there is endless talk about the latest needle in a haystack finder, when what we are facing is a collapse of the market that funds the creation of needles. Netflix caught on. Let’s hope that the people who are funding cancer research and teaching students get a clue soon as well.

I have a deep appreciation for metaphors and analogies that put complicated issues in a language that people can understand, and Mike really does this well. And I find it particularly delicious when someone else helps me understand my own work more clearly.

Kim and I founded Lumen because we “caught on” in the same way that Netflix did. Rather than trying to build a better “OER in a haystack finder,” Lumen’s strategy has been to work with faculty to select, align, enhance, and aggregate individual OER into comprehensive, well-designed collections that people will actually want to adopt (and then continuously improve the individual resources and the collection itself based on student and faculty feedback and relevant learning data). In other words, we’re trying to facilitate OER adoption by creating greater OER leverage. And once a faculty member has adopted OER, then there’s a chance to talk about new pedagogies, student co-creation of knowledge, and the other things we really want to talk about.

I love new ways of thinking about my own work.


Please don’t misread this as an argument for “open textbooks.” This is an argument for leverage. While the collections of OER that are sometimes referred to as open textbooks are large enough to create significant leverage, the language of “textbooks” ties us to the past in ways that subconsciously constrain our beliefs about what we can do with OER. I continue to believe that every time we use the word “textbook” to describe the work we’re doing with OER we paint ourselves a little further into the corner of traditional thinking about teaching and learning resources. This approach might win the battle but it will lose the war.


          1 2