AFT - American Federation of Teachers

Shortcut Navigation:
 
Email ShareThis

Warming up to the machine

By Michael Feldstein

What faculty need to know about adaptive learning.

The phrase “adaptive learning” is an umbrella term that applies to an incredibly broad range of technologies and techniques with very different educational applications. The common thread is that they all involve software that observes some aspect of student performance and adjusts what it presents to each student based on those observations. In other words, all adaptive software tries to mimic some part of what a good teacher does, given that every student has individual needs.

Here are a few examples of adaptive learning in action:

  • A student using a physics program answers quiz questions about angular momentum incorrectly, so the program offers supplemental materials and more practice problems on that topic.
  • A history student answers questions about the War of the Roses correctly the first time, so the program waits an interval of time and then requizzes the student to make sure that she is able to remember the information.
  • A math student makes a mistake with the specific step of factoring polynomials while attempting to solve a polynomial equation, so the program provides him with extra hints and supplemental practice problems on that step.
  • An English-as-a-second-language student uses incorrect subject-verb agreement in several places in her essay, so the program provides a lesson on that topic and asks her to find and correct her mistakes.

In most cases, the software is adapting to details of student performance that would be obvious to any good instructor if she had the time to observe closely enough. Occasionally, there may be some extra bit of cognitive science knowledge built into the program that the average instructor would not know. For example, most teachers probably don’t know the details of how frequently and at what intervals humans should be retested on a memorized fact in order to ensure that fact gets into long-term memory. (And even those teachers who do know generally do not have the time to work one-on-one with students and requiz them appropriately.)

What it’s good for

The simplest way to think about adaptive learning products in their current state is as tutors. Tutors, in the American usage of the word, provide supplemental instruction and coaching to students on a one-on-one basis. They are not expected to know everything that the instructor knows, but they are good at helping ensure the students get the basics right. They might quiz students and give them tips to help them remember key concepts. They might help a student get unstuck on a particular step he hasn’t quite understood. And above all, they help each student figure out exactly where she is doing well and where she still needs help.

Adaptive learning technologies are potentially transformative in that they may be able to change the economics of tutoring. Imagine if every student in your class could have a private tutor, available to him at any time for as long as he needs. Imagine further that these tutors work together to give you a daily report of your whole class—who is doing well, who is struggling on which concepts, and what areas are most difficult for the class as a whole. How could such a capability change the way you teach? What would it let you spend less of your class time doing, and what else could you spend more of your class time doing instead? How might it affect your students’ preparedness and change the kinds of conversations you could have with them? The answers to these questions are certainly different for every discipline and possibly even for every class. The point is that these technologies can open up a world of new possibilities.

What to watch out for

Despite the promise of adaptive technologies, and despite the liberal use of buzz phrases like “big data” and “brain science” by the vendors who create products based on these technologies, adaptive learning systems are not magic. They are tools that should be understood and employed appropriately by skilled educational practitioners. So while they are well worth exploring, there are questions you should ask and issues you should think about before making any big decisions. 

To begin with, if you are thinking about trying an adaptive learning product in your class, it is important for you to understand the ways the software adapts to the students. Before you hire a tutor, you want to know what that tutor can and cannot help your students with. You might even want to watch the tutor work so you can see her skills and limitations. The same is true with adaptive software. There is nothing these packages do that you, as an educator, are not capable of understanding from a pedagogical perspective. If the vendor cannot explain the software’s capabilities in what amounts to commonsense language about teaching and learning—if all you get is techno-babble—then you should think twice about adopting it. There is no reason why you should have to accept a black box as a teaching product.

Can you trust adaptive technology?

One reason you need to understand how it works is so you can decide how much you trust the software to do what it claims it can do. These are your students, and you are turning them over to the care of a tutor. Do you trust the tutor to teach the right concepts and, perhaps more important, not to give false or misleading guidance? How much you trust your adaptive technology depends a lot on what it is supposed to do. A multiple-choice test question that links incorrect answers with supplemental content is easier to make work right than an essay assessment program that attempts to diagnose student writing problems. 

Context also matters. We can tolerate tools that are not perfectly accurate in some cases better than we can in others. Most students learn pretty quickly that a Google search will yield some results that aren’t helpful, and they adjust accordingly. Getting them to understand when to trust a grammar checker and when not to trust it is a lot harder. 

More broadly, it is critical to develop a clear and well-articulated position on which teaching functions the software can fulfill and which it can’t, in order to defend the value of a real college education and the faculty who deliver it. There is a cultural temptation, fed somewhat by eager vendors and a press that tends toward an excess of techno-optimism, to believe that adaptive learning platforms are the future of education and can be full replacements for teacher-facilitated classes. 

The root of the problem is not the adaptive technology itself so much as the belief that a “good” education is entirely quantifiable and therefore manageable by computer. When policies to hold schools accountable for student success get reduced to a handful of all-important metrics, there is danger. When the idea arises that machine-assessed competencies capture everything important that a student should learn in a class, there is danger. In these circumstances, the notion of adaptive learning technologies can be abused as a kind of magic incantation by the reductionists.

The countervailing temptation for faculty, then, is to reject all adaptive learning itself as a fraud and a conspiracy to defund education. That temptation should be resisted. Adaptive technologies can have real value and are not going away. They can free up faculty to spend more time doing what they do best in the classroom—work that is not replicable by a machine. Rejecting these capabilities out-of-hand would risk damaging the credibility of faculty while denying students support that could improve their chances of success. 

The better approach, from both educational and labor perspectives, is to examine each tool on a case-by-case basis with an open mind, insist on demystifying explanations of how it works, embrace the tools that make educational sense, and think hard about how having them could empower you to be a better teacher and provide your students with richer educational experiences. Don’t be content to merely argue that you can’t be replaced by a machine. That’s a losing strategy. The winning strategy is to prove it.


Michael Feldstein is a partner at MindWires Consulting (www.mindwires.com) and founding co-publisher of the e-Literate weblog (www.mfeldstein.com). Previously, he was an assistant director at the SUNY Learning Network, where he was a member of the United University Professions/NYSUT/AFT.

Reprinted from the Winter 2013-14 issue of On Campus.