Categorizing OER based on four purposes

September 17, 2010, [MD]

This part of my thesis is based on a framework that I have gradually developed over the last year and a half. It began with reading Mike Caulfield's blog post Openness as reuse, and openness as transparency, where he contrasted the purposes of MIT OpenCourseWare and CMU's Open Learning Initiative. Interestingly, this was inspired by the same course on open education by David Wiley, that inspired me to begin research Chinese OER. He introduced the notion of OER projects having different purposes, and suggested that the main purpose of OCW was transparency, seeing what other people had done. OLI, he suggested, was more about reuse. About half a year later, I was giving a major talk at my institution about Open Educational Resources, and in thinking through how I could give a systematic overview of the field, I came up with three purposes of OER: Direct use, reuse and transparency (Slides, links, video and audio, this topic starts around minute 36).

As I began researching the Chinese project, reading the literature, and talking with professors about the purposes of that project, I realized I had to add another purpose, which I eventually decided to call transformative production: the changes in the people who produce the materials, caused by the participation in the production and opening of the materials. While I believe this effect is to some extent inherent in all OER projects, the Chinese project was the first one where I saw this as explicitly stated, and put front and center.

I also become very convinced of the value of this analytical framework, when I read a number of Chinese papers that criticized or evaluated the Chinese project, but failed to discuss what the purpose should be. Some would criticize it for not being very suited for independent learning by students, when the initial project plans had never called for material useful to that group. I believe that it is impossible to fulfill all four purposes equally well (although you could probably do two or three), and that by not making an explicit choice, you end up with a project that does nothing very well:

There are many models for developing Open Educational Resources, and this is partly because the goals of the various projects are different. To make this clearer, I propose a typology of Open Educational Resources based on their purpose. When people develop Open Educational Resources, they make many decisions regarding format, scope, organization, licensing and so on, and these are informed by the purpose the resource is to fulfill, as well as technological and organizational limitations. After publication, the resource can be used in many other ways by different users, indeed one of the strengths of the open licenses is to enable this kind of unexpected use and reuse, however the original purpose is still a useful guide. To be clear about the purpose is not only important for the design of the project, but is also a necessity for any rigorous evaluation to take place.

I first introduce the category of transformative production, which I added after beginning to research the Chinese project:

By transformative production, I mean that the process of producing the resource in itself has a transformative effect upon the people involved in the production process. Just as the purpose of writing an essay in school is not to generate a large amount of finished essays, but rather in the effect on the person writing the essay, this category suggests that the purpose of the production of open resources, or the opening of existing resources, is the effect it will have on those involved. This effect is always present even unintentionally, and could be positive – where teachers put more efforts into their teaching, because they know they are being filmed – or negative – when teachers abstain from experimenting in class, because they are afraid of having their failures caught on tape. However, this category covers projects that have this transformative effect as their main goal for the production of open resources. It is different from the three that follow, which all pertain to the resources after they have been produced.

My second category was direct use by independent learners. I realize that independent learners can learn from almost anything, no matter how poorly organized or incomplete, however we would expect something more from material that had been designed with independent learners in mind.

By direct use, I mean that the student can visit the resource and use it to learn independently. This means that the resource would ideally contain all the material needed to learn, ie. be complete. The resource should also be developed for the web, taking advantage of the possibilities offered by interactive quizzes, simulations, games, and other mechanisms. Developing this material might be expensive, and it should be clearly targeted to a specific group. A good example in this category is the Carnegie Mellon Open Learning Initiative courses (see for example Dollar and Steif 2008).

The Open Learning Initiative is a classical example, because it not only provides a complete resource, but it also has interactive exercises, simulations and animations that aid the learner. These resources can be immensely powerful for a large amount of students, but it's extremely expensive, and requires a number of specialized skills to achieve.

By reuse, I mean that the material can be modified, redacted, and integrated with other material. In this case, the student does not directly access the material, but the access is mediated through an intermediary — for example a teacher, or a curriculum developer. In this case, the material needs to be openly licensed, so that the transformation is legal. Material in this category does not need to be complete, or targeted to a specific group, since it will be repurposed. The material in this category is often not organized as an entire course, but as a large collection of small modules. The material should ideally be available in file formats that are easy to modify by the user. A good example in this category is Rice University’s Connexions project, which uses small modules, an open XML file format, obligatory open licensing, a built-in system for derivation and attribution, and a flexible system for quality review to facilitate reuse and the creative building upon other’s work (Baraniuk 2008).

Reuse is kind of the holy grail of the OER movement, and the main reason why we care so much about open licenses. The sad truth is that very little reuse is happening, however, and most institutions are happier producing their own branded material from scratch, than improving on somebody else's material. In Mike's original blog post, he suggested a difference between scripting languages and object oriented modules as an analogy to the difference between packaged learning objects with meta-data, and "just give me your stuff". He elaborated on this in his second blog post:

Transparency (show your code) does promote a certain type of reuse — but it is generally I think reuse of professionals of the same caliber. And this is where the OO vs. scripting language comparison comes in useful — the idea of scripting languages is sort of a single tier — scripters reuse what they learn looking at scripters.

The whole OO idea, when expressed as a business model, was that there are different tiers of user/creators — that the way-smart people make the objects and the less smart (and less paid) people script them together, and this maximizes efficiency.

The everybody is a scripter (which I see as a sort of craft model), and the specialized production OO model (which i see as a manufacturing model) come from two really fundamentally different world views — they intersect in this small place, but at the edges they start to tug at each other.

Once again, I think we need both — the Python Library is a thing of beauty, and allows me to do crazy things with code that I could never do on my own. On the other hand, so much of what I’ve produced of use has come from hacking at spaghetti code copied and pasted from somewhere.

I think there are analogues in open education, even in a single implementation. I might grab the best lecture on Aeration from TU Delft and drop it unedited into my curriculum. I might follow that by reviewing the reading list for that course, and pulling one or two readings I have missed into my own curriculum. But I think even is this case, they are two slightly different activities — in one instance I am essentially a consumer, and in another I’m a co-producer.

In some presentations, I have differentiated between "light reuse" and "heavy reuse". By light reuse, I mean deep-linking to a resource, putting it into another curricular context, but not changing the resource itself. This is what David Wiley did with the original Intro to open ed course, he didn't produce any of the readings himself, he linked to reports, videos and other material that already existed. He didn't modify these resources, but he put them into a new context. Light reuse is much easier to do, and does not require the object you are linking to, to have an open license (as long as it is freely available, ideally without a login).

This is also what we are doing at P2PU currently, and what I have often suggested could be done with the open courses in China... If you criticize them for not being good, or complete enough, why don't you create your own curriculum. There might be 20 courses on intro to economics, why don't you go through and choose the best videos, the best PDFs, the best examples, and link them together into a new curriculum? It's completely legal, and extremely useful to others. This would be similar to the object oriented metaphore, except in this case, we are not asking for the resources to be heavily packaged meta-data labeled learning objects, it's enough that they are on the web, and are direct linkable.

Heavy reuse, on the other hand, means that you transform the materials themselves. You might download them, edit them, take out chunks, put them together with other resources etc. This is much more labor-intensive, and requires that the material is licensed under an open license. The most common form of heavy reuse, is translation.

By transparency/consultation, I mean that the material will not be used directly by learners, nor will it be “reused” or repurposed by intermediaries. Rather, it will be available for people who are interested in learning about how a given class is taught. This could be other teachers, who wish to get inspiration about different ways of teaching the same thing, or students who are planning to choose a major, and would like to know what a given subject entails. It could inspire other teachers, or even provide materials for a comparative curriculum study. This requires material that reflects as closely as possible what actually happens in the classroom, or material that is distributed to students in a normal situation. The OpenCourseWare projects are good examples of this category, and so are the open textbook repositories in India and Indonesia (Ghosh and Das 2006; Hariyanto 2009).

You can read more about the textbook repositories in India and Indonesia on my blog. You could say that there is a lot of overlap in terms of the material that might be made available under reuse, and transparency/consultation. In both cases, we want as much material as possible, don't worry about whether it's useful or not. And this is why the purposes of direct use on the one hand, and reuse and transparency/consultation on the other hand, can never be served well by the same project. Mike Caulfield says:

It would be tempting to say that while separable concepts, the aims of reuse and transparency are so synergistic as to never be at odds. But this isn’t the case. Engineering for reuse takes a certain type of investment that constitutes a drag on transparency efforts. Transparency is most effective when as much is made transparent as possible. The principle behind transparency is that you never know what bit of internal information may be valuable to outsiders. And you shouldn’t really spend too much time worrying about it — get as much open as you can.

Leigh Blackall commented about the difficult of achieving both of these objectives:

At Otago Polytechnic we have been trying to achieve both at the same time, and some may have noticed that I use the term “open educational resources and practices” to encompass that intensive approach. There is a sense urgency in our need to update skills, awareness and policies to a point where we able to offer quality services in open (flexible) education arenas. But as Mike suggests, there is observable drag in doing both.

Although both reuse and transparency/consultation benefits from more available material, they differ in some cases. It is not that important that material in the transparency/consultation category use an open license (although I personally prefer it). Perhaps the most important difference is that the material in the latter category is valuable because it reflects on something else. If I want to learn about how American professors teach, and I watch OCW videos to achieve this, I am not interested in the videos themselves, but in what they reflect (what goes on in the classroom). Thus I can learn very little about how professors at Carnegie Mellon teach classes, because the OLI material is not a reflection of what happens in face-to-face classes.

In a later section, I will use these four categories to discuss how the China Top Level Courses Project, and the MIT OpenCourseWare models are similar and different.

The quotes in this text is from the MA Thesis "The Chinese National Top Level Courses Project: Using Open Educational Resources to Promote Quality in Undergraduate Teaching" by Stian Håklev, University of Toronto 2010.

Stian Håklev September 17, 2010 Toronto, Canada
comments powered by Disqus