Adaptive Courseware: The Good, the Bad, and the Ugly
NAU has received a grant from APLU (the Association of Public Landgrant Universities) to explore Adaptive Courseware with our faculty and students. Before we get into the details, it's a good idea to unpack those words. Courseware is content in software form, rather than a physical, bound, paper textbook. It's also a collection of assessments (tests) of the learner in various machine gradeable (multiple choice and similar) forms. But e-books from the big publishers like Pearson, Wiley, McGraw-Hill, and Cengage have been around for a while and are in fairly general use. It's the adaptive part that's new. Both the course content and the assessments can be adaptive.
So what do we mean by adaptive? The basic idea makes great sense. The idea is that the software monitors the way the students progress through the course material, and how they do on the assessments, records their responses and their path for analytic purposes (noting parts of the course that are causing problems for students, for whatever reason) and presents different content and test questions based on their responses. This makes perfect sense. If a student isn't getting it, or if they are bored silly, it makes no sense to keep throwing more of the same stuff at them. It makes much more sense to divert the struggling student into remedial or preparatory content, and to present deeper, more sophisticated content to the student who isn't being challenged.
What are the reasons students have problems with the content? The three most obvious ones are that 1) the content is unclear, 2) the content is not engaging, or 3) the content is not appropriate to the student's level of prior knowledge. With data, we can fix these things and make the course better, with the ultimate goal of aiding the learner.
Why is this, potentially, a great idea? Retention of students is better than losing them and recruiting new ones. Making sure that students are ready for the courses they are enrolled in increases student success in a meaningful way, makes professors happier, and ensures that the reputation of the institution improves.
How does this transform education? As content is personalized, it breaks the traditional lock-step approach, where (if we're lucky) the middle 50% are getting it, the bottom 25% are failing, and the top 25% are bored. It allows struggling students to get the background information they need to succeed, and it allows the advanced student to finish early or go farther, and to have that deeper level of mastery get recognized. It allows everyone to proceed at their own pace, and to spend as much time as they need on the parts that are challenging, and to go deeper into material that interests them.
Well, that all sounds great. What's the catch? That's next...
For our first round of product reviews, we're looking at four vendors: Acrobatiq, Knewton, Cogbooks, and Learning Objects Difference Engine (a division of Cengage). What we have quickly realized is that each of these tools has a very different user interface, which goes against the efforts we've been making to create a more consistent look and feel across all NAU courses. This makes it more difficult for students to navigate because every course might be very different in its layout and design. It also relegates our learning management system, Blackboard, to the role of a portal. Students log into Blackboard, find the course, click on a link, and then leave the LMS and land on one of many possible courseware sites. With lots of integration work, we can get grades back into the LMS, but that's about it. Nothing lives in the LMS.
We have also realized that adopting these tools results in a considerable loss of intellectual freedom over content and delivery of a course, because the content is deeply intertwined with the adaptive engine. In order to get good analytic data on learner behavior, content needs to be thoughtfully tagged, and progress needs to be carefully tracked, and this makes authoring content more challenging than just knowing the subject matter. In most cases, authoring is a job for professionals, not content experts. From the vendor presentations, it's not clear what the role of the instructor is with these tools, and therefore it's hard to see where our faculty experts can add value or personalize the content.
Difficulty customizing content raises another question. How does one school differentiate itself from another if they are both using the same product, and the product is not very customizable? How does NAU compete against a school with lower tuition costs, for example? How does an NAU instructor integrate content on life at elevation to a Biology course; something that is both relevant and interesting when you live at 7000 ft? When students can shop around for courses, why would they pick our version?
Another common objection is that students who don't get it have more work to do, or that students who are doing well get more work piled on. This may seem reasonable, but it is a very foreign concept for most students, and the students will resist it unless the rewards are tangible. We haven't transformed the rest of the educational process, so those rewards are not clear at present.
The final issue is that, beyond the introductory level, there isn't (and may never be) a lot of adaptive courseware content, because it's a niche market. There is also not, at present, much content that isn't machine gradeable. That means that liberal studies and the humanities are not going to be well-served in the near term. Therefore, this is not a complete solution.
The APLU grant is only seed money. Purchasing adaptive courseware is a new financial burden that will fall either on students or universities. It's pretty clear that this is the future and, if they get it right, the potential is great. But it's also clear that we're moving towards a world where people are taught by machines, and the human factor is being pushed to the side. Are we doing this because it's a better way, or because it reduces costs? I don't know about you, but a computer telling me, "Great Job, Larry" isn't very motivating.