In Spring 2016, faculty, support staff and administrators at Oregon State University met to candidly share their experiences with adaptive learning technology.1 I shared two different videos from the event at EdSurge in this article and highlighted comments on vendors over-promising here at e-Literate. This time I’d like to highlight part of a panel discussion where a faculty member relates her experiences – what worked and what didn’t work – when using adaptive learning tools.
Kathryn Becker-Blease has taught Intro to Psychology, a large lower-division lecture course, using both traditional quizzing and with adaptive quizzing with the help of Macmillan’s LearningCurve. In this part of the panel discussion Susana Rivera-Mills, Vice Provost and Dean of Undergraduate Studies, asks Becker-Blease about her experiences based on research and teaching.
It is worth noting the specific application in question – the academic discipline, course format and size, tool usage – as adaptive learning can mean different things in different contexts.
With that in mind, however, I noted in the panel discussion in response that the challenge of working with lower-performing students based on their study habits and support needs reminded me of the experience at Essex County College that we shared last year. In that case, a remedial math course was redesigned using adaptive learning tools – McGraw-Hill’s ALEKS – but with an important focus on self-regulated learning. Yes, students worked in the adaptive learning platform, but they also had class time devoted to supporting them with their study and work habits. Learning how to learn, which is very important for students that do not have a history of academic success.
Most academic initiatives require holistic solutions, particularly with the complex task of reducing achievement gaps.
- Disclosure: Our e-Literate TV series of video case studies and explainer videos is funded by a grant from the Bill & Melinda Gates Foundation. [↩]
Hey Phil,
I couldn’t agree more: adaptive learning and self-regulated learning go hand-in-hand. I’m sure the OSU psyc faculty are familiar with the pioneering work of Barry Zimmerman in SRL, but those of us in higher ed IT could benefit from understanding Zimmerman’s research, and its implications. A few years ago, the Chronicle did a nice piece on Zimmerman’s research at CUNY:
“Struggling Students Can Improve by Studying Themselves” by David Glenn (2/7/10)
http://www.chronicle.com/article/Struggling-Students-Can/64004
Unfortunately, it’s behind the CHE’s pay wall, but among other things, Zimmerman identified two keys for SRL to “work” for struggling students: 1) they had to own their deficiencies, instead of blaming others (e.g., the prof doesn’t like me, or the textbook is too hard), and 2) feedback had to be specific on demonstrated weaknesses vs. generic, general aphorisms (e.g., “study harder”). The first is tough to break through because by the time students come to higher ed, they may be well-versed in looking externally to explain or justify their poor performance. But this is where the social or peer-comparison of nudge analytics might help. When students see how they compare to peers, there’s at least an opportunity for them to see things as they really are. Some will, some won’t.
The second area of “specific feedback” is where I think adaptive learning could be most powerful, but also difficult to scale. At its best, adaptive learning should help students uncover or display what they are weak at. Maybe it’s solving equations with two variables or verb tense. But then the adaptive learning experience should tutor and re-assess them in their specific weaknesses. Rinse and repeat. To me, I think Khan Academy is one of the best examples of this. I was an English major, but only ok in math, due in part to going to three different high schools following a cross-country move. However, I was recently able to get through my two doctoral level stats courses with the help of a good, patient instructor and Khan Academy. By this point, I’d run out of excuses about our move, because that which I was most interested in (data mining student and faculty use of the LMS) was behind the wall of statistics that loomed large and impenetrable. I won’t say I’m a whiz-bang data scientist, but I can read a journal article and present my research in ways I just couldn’t before.
I know some faculty dismiss KA for occasional errors or a perception they could do better. Maybe they can or can’t. But whether we build or buy an adaptive learning platform, we would do well to keep in mind the holistic approach you call for here. If we want to help students learn how to learn, we could do worse than diving deeply into self-regulated learning.
Thx,
John
Great points, John. Side note: The Essex County College program design was heavily based on John Hudesman’s input; John was co-researcher with Zimmerman. See:
http://e-literate.tv/e3-s20/