The containers for scholarly information are evolving, and the impact is felt everywhere. The containers have to change, because the traditional publishing modes create too much friction. Traditional textbooks cost too much money, and don’t take advantage of adaptive learning. Scholarly publication cycles are too long and too slow for the pace of researchers. Formats imposed in the age of print have proved to be sub-optimal for the reproducibility of science. Intellectual property regimes too frequently inhibit what you can do with materials needed for teaching, and are too complex for everyday use. Key contributions to the advancement of knowledge (tenure-worthy contributions) are swimming happily outside the journal article and monograph pool. Discovery technologies have us thinking about the presentation of knowledge in new ways (data! simulations!), and learning science pushes that envelope further as we endeavor to share knowledge in efficient ways for learners at all levels.
A couple years ago, remarks at a “libraries and MOOCs” event by Christian Terwiesch (Wharton School professor and co-director of the Mack Institute of Innovation Management, among other things) provoked my thinking about libraries and the information ecosystem. His comments, paraphrased years later (and apologies if I don’t get it just right) made three important points:
Teaching and learning requires an ecosystem of quality information
Academic libraries developed a good ecosystem for quality information in the world of print resources and face to face university teaching
Academic libraries need to be deeply involved in new ecosystem development to keep step with changes in higher ed
In this post, I’m focusing my attention on information and information services that support teaching & learning. (As opposed to thinking about research collections, preservation or any of the other aspects of the ecosystem.) And I’m thinking about MOOCs and new business models for global education.
If we think inside the box of current constraints and traditional activities, when we approach new forms of teaching like MOOCs, we come up with services like checking and clearing copyrights, or the probably unsustainable activity of faculty or librarians searching for open educational resources (OER) that match the MOOC needs. Why do we approach it this way? Because the traditional ecosystem around a course involved the set reading list.
How do you get yourself outside the box, outside the frame of how we did/do things in a world of mostly local teaching and locally owned print materials?
I got some help thinking about this in a recent webinar on personalized education, particularly from the presentation by Drew Paulin, Manager, Learning Design and Innovation, Sauder School of Business, University of British Columbia. I’m not going to re-create his very insightful remarks – instead, drawing heavily on his remarks, this is my take on how they apply to the problem space I’m exploring.
The problem space looks like this: In the F2F/print resources world, we have been used to Fixed Content – the reading list, the reserves, the recommended readings. We optimized for the F2F/print context – fixed content results in high quality resources, with high relevance and applicability to the course.
There were no real down sides (or so we thought). Access issues didn’t arise except in the single book–multiple readers situation, which we solved by putting materials on reserve.
But once we leave the local campus and go into a more porous education system (MOOCs, lifelong learning, etc) we see the disadvantages of the Fixed Content paradigm.
In open education environments, you immediately confront what I call the student variability problem (wide differences in prior knowledge and skill levels among students). Paulin and others also refer to the anomalous state of knowledge (you don’t know what you don’t know). Both matter.
The student variability problem also existed in F2F classrooms–and it was exacerbated by fixed content (some materials pitched at a level too difficult for some students). Coping with variable skills often involved: narrow university admissions for certain types of students, pre-requisites for certain courses, faculty office hours for struggling students, and TA-led review sessions. And, we accepted that some students wouldn’t learn as much.
But the solutions of the F2F world for student variability don’t work so well for the more porous education models that will emerge with MOOC platforms and changed business models. The way Paulin put it, you have to get the resources to perform for you. He used terms I didn’t fully grok, but the ideas I came away with include getting the resources to solve problems via:
- personalization for prior knowledge and skill levels, and
- customization for student goals (familiarity? mastery? certificate?).
In the case of customization, we recognize that people need to learn for different kinds of reasons — a researcher who will perform statistical research needs different levels of mastery than an administrator who wants better skills to evaluate research and be more evidence-based in practice.
Maybe the highly fenced-in, pre-selected single set of resources is not going to be the answer for the course of the future…
If we turn to OER for the answer, we find lots of good things. The access issue (openly available stuff, not just locally available for some tiny set of students) is addressed. And there are abundant resources available via OER that can solve the personalization issue (what I can understand based on my prior knowledge) and even to some extent the customization issue (fits my goals) — but!
OER exist as a giant mosh pit and it’s hard to find stuff. How do I, a student, go after materials appropriate to my prior skills and knowledge, and my learning goals? What should I read first? We lose the pre-selection for high quality and applicability. We lose efficiency for the student. There’s going to be a lot of labor involved, it seems, to get the student matched up with the right resources in order to efficiently learn the course materials. Students themselves are not good at finding appropriate materials much of the time because of the anomalous knowledge problem – the don’t know what they don’t know, so they don’t know what to look for. If we think we’ll need experts to find materials to address personalization and customization, that seems like an overwhelming amount of expert intervention when we are talking about thousands of students. But if we don’t take up the particular advantages of OER — for personalized content, for customized content — then we are simply using the old paradigm of fixed content to teach to the middle. In a new environment that has new affordances, surely we can do better than tread water?!
The Fixed Content approach has its advantages and disadvantages; the OER mosh pit also has pro and cons. What’s the new optimal ecosystem?
How should we solve these problems:
- efficiency – getting users to the right materials (for them!) quickly
- high quality resources
- high applicability resources, just the right resources for the course topics
- access – stuff people are allowed to use or can access based on their affiliations and location
- personalization – aimed at my prior knowledge and skills, helps me forward from where I am
- customization – is geared toward my own goals for learning
Thinking about the problem space in this way gets me to imagine how linked data and computational methods (rather than unscalable human interventions) can bring the right materials to users. Metadata, crowd sourcing, linked data, recommendation systems and feedback loops, connecting user profiles (“I live in Bangalore”) with access metadata, and so many other options are available to us. If users contributed information about themselves and their affiliations, it could be linked to data already existing in global library databases, and layered with ways to improve the accuracy of initial recommendations.