Some Lessons Learned Supporting OER Adoption

The tl;dr:

• Supporting effective OER adoption at scale has its problems.
• Many of these problems have openly licensed solutions.
• Sometimes it makes sense to deploy these solutions yourself; sometimes it makes more sense to work with a partner.

Background and Some Problems

To put it in a depressingly small nutshell, I spent the first decade or so of my career creating open licenses to make the sharing of OER legally possible, traveling the world talking to people about why they might want to place an open license on their educational materials and other creative works, experimenting with different open pedagogies in my own teaching, and conducting empirical research about the impacts of OER adoption on outcomes for students, faculty, and institutions.

Upon reflection several years ago, I came to see that despite all my efforts I was making the classic Field of Dreams mistake (“if we build it, they will come”) by assuming that “if OER exist in a faculty’s discipline and research shows them to be effective in supporting learning, faculty will adopt.” This turned out to be true only for a very narrow range of faculty – generally those who were previously innovation-minded (those same seven or so faculty on each campus that eagerly try every new thing). If OER adoption were to become widespread among the majority of faculty, it became clear that someone would need to do something more than create OER, post it on a website, and give conference talks about it. This is why we started writing grants focused exclusively on supporting OER adoption rather than on funding new OER creation. After all, there are over a billion CC-licensed works now – not everything we need, certainly – but enough that it felt like someone ought to be focused on helping faculty use what is there.

Over the last several years my fellow travelers at Lumen and I have learned a lot of painful lessons about supporting OER adoption among faculty. (In fact, I think it’s safe to say that we know more about ineffective OER adoption techniques than anyone!) However, working through and responding to these early challenges together with amazing collaborators at a wide range of 2-year and 4-year institutions around the country, we’ve learned something about how to support effective OER adoptions, too.

Perhaps unsurprisingly, some our earliest and most painful experiences with supporting effective OER adoption involved the campus learning management system (LMS). From the very beginning, we have felt strongly that learners should work with OER directly inside their LMS. Sending learners outside the LMS to a secondary system is bad for a range of reasons, including the fact that it creates extraneous cognitive load by forcing students to dedicate some degree of attention and learning to the second system’s novel interface, meaning this attention and effort is unavailable for the purpose of learning biology or history.

As we worked on our first OER adoption support grant (a Wave I Next Generation Learning Challenges (NGLC) grant) we began by helping faculty cut and paste OER directly inside their LMSs. This, of course, turned out to be a complete nightmare. We manually rebuilt each course half a dozen times or more, in Blackboard, Sakai, Canvas, and other systems. Given the differences in Common Cartridge implementations across systems, a clean build was often faster and always less error-prone than a CC import and cleanup, especially when you never knew what was going to break where.

It also became clear rather quickly that this approach essentially threw OER to the wind, making it “difficult” to make timely, critical improvements. Say, as a purely hypothetical example, that a member of a faculty team accidentally included a copyrighted resource in a collection of OER. Once those (supposed) OER were copied and pasted in half a dozen LMSs at a dozen institutions, and they began the local course cloning process for faculty, how do you put that genie back in the bottle? (I’m asking for a friend.) The flip side of this problem is that there’s also no way for positive changes that faculty make to OER – revisions and remixes that might benefit everyone – to propagate back up to the community. Everything is trapped inside the LMS.

And because you’ve trapped your OER inside the LMS, the day after class ends all your students lose access to their OER. Think about that for a minute. Philosophically, the idea that the LMS deny students access to OER was the most annoying of all the problems.

And then there were the attribution problems. The most enlightened LMSs provided an option for you to CC license your course materials. Unfortunately, this was a course level setting that resulted in a single CC license icon / attribution statement being shown in the footer of every page. Of course, this approach to attribution only works if every OER in the course comes from the same source and was published under the same license. Having the wrong attribution – or license – listed for a resource is highly problematic. Licensing is critically important and needs to be handled correctly.

To solve this problem we turned off course level CC licensing tools, created our own attribution generator that spit out HTML, and started appending the correct attributions to the bottom of every page throughout each course. It turns out that while faculty hate plagiarism, they don’t always immediately “get” the attribution requirement of open licenses. For whatever reason, sometimes they just deleted the attribution statements from the pages in their courses – so that the resources weren’t attributed at all. Now combine this problem (missing attribution) with the previous one (throwing OER to the wind), where a resource that is missing attribution is propagated across course shells for all the faculty who teach a course on a campus…

Feel like hitting your head against a brick wall yet? We did.

Working Towards Answers

With project funding from the Shuttleworth Foundation, and in consultation with our institutional partners, we began working on an alternative approach that would allow us to have our cake and eat it, too. The idea was to extend the open source WordPress software in ways that would make it an extremely lightweight and easy to use OER management and integration platform – that would still embed OER directly within the LMS for learners.

WordPress was a great choice for many reasons. Its editing tools are so easy to use that your grandmother probably has a blog (by some estimates over 20% of all sites on the web run on WordPress). This would keep faculty training requirements as low as they can be kept. WordPress automatically versions updates to each page, so that you can easily (and programatically) see which resources have been revised / remixed and which have not, making it easy to flag improvements for review and possible upstream adoption. And of course WordPress pages have stable URLs that will persist after a course ends, so that students can have ongoing access to resources. The open source Pressbooks plugin makes some revise / remix tasks simpler (e.g., drag and drop changing of chapter order) and makes it significantly easier to export resources in formats that humans can use, like PDF and ePub. And since it’s also open source, the Lumen dev team have been able to make several contributions to the Pressbooks codebase.

Starting from here, we built a new, open source CC licensing plugin for WordPress with a focus on creating and managing multiple attributions per page (because we wanted to have strong, native support for revise and remix). In addition to capturing and managing licensing information as page metadata (instead of directly within the content of the page as we had done previously), this tool also lets you roll up, view, and manage all the attributions in a course from a single page in the WordPress admin.

Then we created a new, open source LTI plugin and a new, open source Thin Common Cartridge plugin for WordPress. The Learning Tools Interoperability (LTI) plugin allows you to embed content directly within the LMS so learners see content where they expect it to be and there’s no interruption in the learning experience. The overwhelming majority of learners can’t tell the content hasn’t been copied and pasted directly into their LMS (except that when you can control the styling, your LTI-ed content often looks quite a bit nicer than the surrounding LMS). The Thin CC plugin allows you to create a Common Cartridge comprised exclusively of LTI links to all the resources you’re managing back in WordPress. This way, with one export / import, you automatically create individual LTI links to all your WordPress-based OER within the LMS (one link in the LMS for every WordPress page). Then faculty can use LMS tools to rearrange, remove, or organize links in modules or folders, depending on their system’s UI paradigm, and they can jump to the Lumen platform to do more significant revising and remixing (adding or changing photos, removing, adding, or rewriting examples, etc.).

(We’re still working on a new, open source outcomes alignment plugin for WordPress. This plugin currently works, allowing you to align resources in WordPress with learning outcomes, but the authoring experience still leaves something to be desired. The same is true of the Open Assessments tool which we use a lot and are active contributors to.)

By creating, extending, and integrating these open source tools, we’ve been able to overcome these and many other problems we had faced over and over again. The answer was to move the managing, attributing, revising, and remixing work outside the LMS, while still allowing learners to use OER within the LMS. For us, this has been classic open source work as characterized by Eric Raymond in The Cathedral and the Bazaar – “Every good work of software starts by scratching a developer’s personal itch.” And this approach has been very well received by students and faculty alike.

If you’ve been tracking carefully, you may be a little confused at this point. OER are freely available and come with 5R permissions. Lumen has spent a lot of time, effort, and money creating an OER management and integration platform that solves many of the most common OER adoption problems, which is also free and open source. This may lead you to ask – how does Lumen continue to exist? Why on earth would anyone ever pay to work with Lumen? There are many answers to these questions; let me briefly discuss the ones most germane to the lessons learned I discussed above.

Of Red Hat and Reclaim

First things first – some people will never pay to work with Lumen – and that’s fine. They’ll install and configure our open source software and load it up with open content from all over the world (probably some of the same OER we’ve worked with faculty to curate and aggregate). They’ll double check that the licensing and attribution are correct on the OER they’re using. They’ll provide training and support to their faculty, they’ll host and provide technical support for the platform, and (hopefully) they’ll do their own analytics work and effectiveness research. Power to these people! That independent, self-supporting spirit is a big part of what open is all about! Hopefully, many of them will contribute improvements to the components of the Lumen platform, just as we contribute to other open source projects.

Institutions that aren’t in a position to provide all these services themselves, and institutions for whom it costs less to partner with Lumen than it would to provide these services themselves, will likely choose to partner with Lumen. In this regard, we’re a bit like Red Hat. Or, if you prefer an example closer to home, Reclaim Hosting. They curate and extend open source software with their Domain of One’s Own approach to owning and managing your digital presence. Any institution could go grab the same open source software, do the integration work, and run their own, similar project without partnering with Reclaim. But it’s generally easier and less expensive (not to mention more fun) to just work with Reclaim than it is to recreate their efforts. Institutions partner with Lumen because we provide faculty training and support, checks of OER licensing and attribution, hosting and technical support for our platform, and analytics and effectiveness research – as well as other services like strategic and change management consulting for academic leadership. And because that kind of work – the work specific to supporting effective OER adoption – is all we do, we can often do it both more affordably and more effectively than a single institution can.

Institutions also choose to partner with Lumen simply because we have more experience with OER adoption than they do. We’ve provided direct support for programmatic OER adoptions at 80-ish campuses around the country and have seen a wide range of more and less effective models. We’ve also created a highly effective model for developing, running, and supporting fully OER-based degree programs (we currently support over a dozen of these, but that number is going to change dramatically next week – stay tuned). And perhaps most importantly, from that very first NGLC grant down until today, we’ve connected faculty and institutions into a broad network of other faculty and institutions that are going through the same OER adoption process they are, creating opportunities for collaboration and networking across institutional boundaries.

This list isn’t exhaustive, but I think it hits the high points relative to the problems and solutions I described above.