Tag Archives: Adaptive Learning


Why Students Don’t Like School — and What Adaptive Learning Can Do About It (Part 1)

Ask students why they don’t like school, and you’ll get several answers: it’s “hard,” “boring,” “disconnected from reality” or “only for smart people.” The real answer is of course more complex than any of these responses would suggest. To get a deeper understanding of the matter, I recently read one man’s investigation: Daniel T. Willingham’s Why Don’t Students Like School: A Cognitive Scientist Answers Questions About How the Mind Works and What It Means for the Classroom.

As I was reading, I noticed that most of the real reasons Willingham argues that students don’t like school can be eliminated or reduced through continuous adaptive learning technology. Here’s how:

1. Work pitched at the wrong level.

Willingham begins his book by debunking some conventional notions about what exactly the human mind is designed to do: “Contrary to popular belief, the brain is not designed for thinking. It’s designed to save you from having to think, because the brain is actually not very good at thinking. Thinking is slow and unreliable.” Willingham indicates, however, that “people enjoy mental work if it is successful.” Hence the popularity of crossword puzzles, sudoku games, and brain teasers. What makes mental work enjoyable? The snap of discovery, the sudden moment of insight. Mental work becomes fun and even entertaining if it consistently yields such moments.

When students complain that school is boring, what they’re probably saying is that it’s either too hard or too easy. The challenge is to get the balance just right: too easy and there’s no satisfaction; too hard, and students will invest effort only to feel frustrated and lose focus. Thus, the key to maintaining student engagement is to escalate the difficulty of the work incrementally, so that students receive a constant stream of questions targeted at the precise level at which thinking and real engagement are likely to occur. Continuous adaptive learning can provide this by determining a student’s ability and “serving up” questions at just the right level.

Of course real life doesn’t happen this way–you don’t get a series of challenges perfectly calibrated to your level, so that every exertion leads to maximum satisfaction; the hope is, however, that adaptive technology can be harnessed so that students engage productively with schoolwork and are therefore better equipped to tackle “imperfect” challenges in the real world. Think of it this way: an adaptive learning system is like a superior mental work-out machine that leaves you ready to scale intellectual cliffs and undertake marathons of critical thought.

2. Not enough opportunities for engagement.

The above paragraphs are premised on the fact that students have enough problems to solve in the first place. If students are only given lectures with minimal opportunity to exercise their cognitive muscles, they will obviously be less engaged.

These “cognitive-work” opportunities are inherent to adaptive learning systems. After all, a continuous adaptive system is based on the idea that what you see going forward depends on your previous activity and performance. In other words, it’s practically impossible to design a continuous adaptive learning system that doesn’t give students a chance to “show what they know” in a fairly constant way. Thus, keeping students mentally active throughout a classroom session is a fundamental challenge that adaptive learning solves.

3. Slow feedback.

The above point — that students need to be active to be engaged — seems an obvious one, but consider from a teacher’s perspective how difficult it is to build problem-solving into every single lesson. The trouble with student problem-solving is that it generally requires feedback of some sort (grading, evaluation, commentary) and good feedback takes time to generate. In this way, the administrative aspect of many productive class activities can make the work for teachers spiral out of control.

As far as evaluation is concerned, adaptive learning can efficiently provide high-quality student feedback, reducing administrative burden on teachers and enhancing student engagement. Whether it’s multiple choice, free response, or even an essay that’s submitted, a continuous adaptive learning system can process student work and deliver personalized assessment. (For more on how adaptive learning works with material as subjective as English composition, check out my post on adaptive learning for soft subjects.) Most importantly, the feedback provided by an adaptive learning engine (designed for continuous as opposed to single-point adaptivity) can be instantaneous or near-instantaneous. This enhances student engagement because students are less likely to lose focus if feedback is immediate and they can quickly self-correct. The result is pacing conducive to risk-taking, experimentation, iterative development, and rapid learning.

4. Lack of background knowledge.

Anyone who’s ever had trouble with the reading comp section on any standardized test (think GMAT or GRE) understands the soporific effect of subjects like the “electromagnetic spectrum” or “sessile organisms.” However, smart test-takers know that the subject itself is supposed to be irrelevant; critical reasoning ability is what’s being tested. For the most part, this isn’t a problem on standardized tests; the obscurity of the content is a neutralizing factor that makes the exam more fair. With schoolwork, however, the subject matter used to impart analytical and creative skills can put students on unequal ground and disadvantage students who have weak background knowledge or have simply not been exposed to certain vocabulary or jargon: “Research from cognitive science has shown that the sorts of skills that teachers want for students–such as the ability to analyze and to think critically–require extensive factual knowledge.” In this way, Willingham asserts, “factual knowledge must precede skill.”

Think of it this way. If you have no experience in economics, you can still read The Economist and get something out of it; but a trained economist will be able to read the magazine much faster, extract the important details, ask intelligent questions, and put the knowledge to work more quickly. Not because he’s a more gifted critical thinker but simply because he’s developed an intuition for the material due to deep functional exposure.

What does this have to do with adaptive learning?

A) A continuous adaptive learning system can provide a scaffolding of hints (definitions, encyclopedic knowledge, formulas) to help level the playing field for those students who have had less exposure to culture, world events, and certain types of vocabulary and jargon. This will allow students to absorb the background knowledge seamlessly and focus on the analytical and creative aspects of any exercise designed to improve their skills in those areas.

B) Adaptive learning can help students learn more efficiently and effectively and in the process, expose students to a range of material in a shorter amount of time (this is related to my point below). Depth and range of exposure can improve a student’s “chunking” ability. Even the simple act of locating a subject in relation to other subjects (an option afforded only by scope of exposure) can make something “click” for many students.

C) Willingham defines “chunking” as “the phenomenon of tying together separate pieces of information from the environment.” Students are thus able to absorb complex knowledge by breaking it down into smaller, manageable chunks. The same goes for problem solving: students tackle complex problems by perceiving them as a series of manageable steps. Adaptive learning can determine what a student needs to grasp before he can have this kind of insight–whether it’s background knowledge, a highlighting of structural qualities in the information, or a certain breadth of range, or a combination of all these elements. In this way, students can be guided toward making those “chunking insights” themselves.

To achieve benefit #3, it is especially important to develop “continuous” as opposed to “single-point” adaptivity.

Stay tuned for more ways that adaptive learning is changing the way students think about school!

Adaptive Learning for “Soft” Subjects: Can Technology Encourage Creativity?

DSCN5354It’s easy to see how online adaptive learning can be used to improve the teaching of quantitative subjects such as math, a subject we perceive to be defined by drilling, discipline, “right/wrong” answers, and skills which build neatly upon one another. Subjects such as English, art history, and music seem to lend themselves less naturally to adaptive technology, whether it’s the “expressive” nature of the subject or the complexity of assessment (i.e. paper-grading) that lies at the heart of our discomfort with the idea.

The requirements for successful powering of “soft” subjects are there if you take a careful look: “soft” subjects, like hard, can be broken into component skills, the mastery of which can be easily assessed. And because the very nature of many “soft” subjects (take Writing for example) is that the technical aspect (grammar, for instance) is inherently linked to the expressive, adaptive technology can ultimately be harnessed in the service of creativity and expression as well as mere efficiency.

Adaptive learning in English composition

Let’s take as an example English composition, a required course at most universities. Most teachers want their students to walk away having mastered the revision and research process. What often prevents the achievement of these goals is the burden of administrative responsibilities and the stresses of classroom management (hand-grading 20 student papers, keeping meticulous score of absences, ushering 20 students through a mandated revision cycle). Also consider that most university comp instructors are paid by the course at as low as $1700 for a 15 week class with 20 students. In such a setting, high-minded goals such as “enabling each student to craft his own style of expression” seem disconnected from reality.

Imagine, however, a blended learning solution: an online segment guides students through a lesson on sentence fragments and run-on’s, serving up targeted exercises afterwards, and leaving class time available for, say, a discussion of Stanley Fish’s How to Write a Sentence: And How to Read One. Such a solution would combine the rigor of drilling with activities that take full advantage of the classroom environment—and ultimately strengthen the connection between technical mastery (the “work” of writing) and creativity of expression.

Improvements in assessment

Let’s take a closer look at a sample composition grading rubric, at the areas which dictate instructor evaluation of student papers: “style,” “clarity of purpose,” “organization,” “grammar,” etc. Adaptive learning can provide personalized instruction for students across all these areas. All that’s required is a logical breakdown of the subject at hand and a method of evaluating proficiency. An area such as “grammar” breaks down into component areas (parts of speech, sentence fragments, verbs, etc), the knowledge of which can easily be evaluated through multiple choice questions. Even “style,” a more nebulous area, breaks down if you think about it: emphasis, parallelism, periodic sentences, etc.

As a former composition professor, I identified hundreds of areas of weakness across student papers (“lack of support,” “wordiness,” “tone shift”). Frustrated that I was repeating nuggets of advice, I started to code certain comments “A,” “B,” “C,” and so on, so that each letter corresponded with academic content related to the identified weakness. Adaptive learning can facilitate this sort of grading on a much larger and more efficient scale, allowing teachers to generate quality feedback that links to personalized programs for each student.

New opportunities for enrichment and assessment

The opportunities for enrichment are endless. A student who is gifted at composing sentences, for instance, might be exposed to content on Proust or Nabokov. He or she could navigate deeply, unlocking worlds of content and instruction, and developing a sense of pride in his/her mastery (much in the way video game players are galvanized with each new level they attain). As students gain confidence, they could assume online “mentorships” where they are encouraged to “teach” others (through peer review and other exercises), increasing the instructional “surface area” of the classroom, building a sense of community, and multiplying the effects of every learning moment.

In the process of all this, online adaptive learning can also neutralize elements of school that detract from learning. Shy and disadvantaged students can gain confidence and experiment with language, without the stresses of face-to-face interaction. In this way, adaptive learning encourages a spirit of risk and facilitates honesty in intellectual exchange.

Revealing the connection between discipline and creativity

The hard work of teaching is showing students how to transform themselves, how, ultimately, to teach themselves. Getting students to the point where they experience the first self-transformation is often difficult; adaptive learning can speed up this process, generating a whole series of transformations within a semester. In composition, for example, students need to see, to feel the connection between discipline and creativity. They need to grasp how the “nitty gritty” and often uncomfortable aspect of mastering grammar and tinkering with sentences will allow them to communicate more freely and even experience thoughts they were unable to before. The targeted practice and quality feedback facilitated by adaptive learning – accompanied by spirited class discussion – will make the class experience richer and generate transformations more efficiently.

And so it is not merely that students will improve their grammar and organization skills, at the same time becoming more comfortable with creative expression – it’s that they will begin to see how deeply connected the technical and the expressive are (which, of course, is the whole point of writing).

Transforming the classroom

Adaptive learning has the potential to make “soft” subject learning not just faster but better. Ultimately, it’s not just about “efficiency,” “comprehensiveness,” “atomic granularity,” or “leveling the playing field”; it’s about powering and transforming the classroom altogether:

  1. By providing precise and individualized instruction in skills areas for which there are “right/wrong” answers
  2. By providing precise and individualized instruction in more subjective areas of “expression”
  3. By strengthening the connection between #1 and #2
  4. By fostering a more inclusive environment
  5. By reducing administrative burden
  6. By encouraging a framework of constant improvement
  7. By affording classroom activities greater flexibility, leaving more time for creative assignments, debate and discussion, group work, and research projects

The Chronicle of Higher Ed: Adaptive Learning "Gives Students More Control"

Photo © Chronicle of Higher Education

In an article released today titled “The Rise of Learning Machines,” The Chronicle of Higher Education profiles Knewton’s partnership with Arizona State University. (This fall, Knewton’s Adaptive Learning Platform™ will power ASU Online’s College Algebra and College Math courses).

The article discusses some key features of Knewton’s Adaptive Learning Platform™ (“programs like Knewton can pace an entire math course using sophisticated tracking of skill development, instant feedback, and help levels based on mastery of concepts”), and calls attention to the underlying problems we’re working hard to address.

As Josh Fischman, the article’s author, writes:

The approach [adaptive learning] is attractive because of some unattractive numbers. Just 22 percent of students in the United States complete an associate degree within three years of starting, and only 57 percent complete a bachelor’s degree within six years, according to the Education Department. Such statistics, along with the large numbers of students who need remedial courses and drop out, drive the appeal of software that offers individualized attention to get students through basic math and other courses that are essential to college success. “These are high-risk, low-socioeconomic-status students—exactly the kind we have to reach out to,” says Mr. Regier [dean of ASU online].

Read more about the advantages of and questions remaining about adaptive learning (along with the response of ASU students and professors to Knewton’s course interface) here, and let us know your reactions in the comments!


What is Adaptive Learning?

David Kuntz is the Vice President of Research at Knewton

If you’ve spent any time in the field of educational technology, you may have heard the term “adaptive learning,” or one of its many aliases: adaptive instruction, adaptive hypermedia, computer-based learning, intelligent tutoring systems, computer-based pedagogical agents…

If you’re like most people, however, the precise definition of the term(s) probably still eludes you. So the question remains: What is adaptive learning?

At the most basic level, adaptive learning is the notion that computers can improve educational outcomes. However, until recently, most adaptive learning approaches have failed to realize this promise. Early attempts were often small-scale, focusing on a limited number of students or area of interest. Most utilized systems with only the most basic kind of adaptivity (eg. “Present Question A—collect the answer—if correct, branch to Question B, if incorrect, branch to Question C.”)

Just as adaptive learning’s name evolved over time, however, so has its potential to revolutionize education.
Continue reading