The Quality of Homework is Not Weighed

by Robert Pondiscio
September 12th, 2011

The perennial debate over whether U.S. students get too much homework or too little misses the point. The question is not quantity, but the quality of the assignments, says science writer Annie Murphy Paul. Writing in the New York Times, Paul says what should matter is how effectively homework assignments advance learning. Neuroscientists, cognitive scientists and educational psychologists “have made a series of remarkable discoveries about how the human brain learns,” she writes. “They have founded a new discipline, known as Mind, Brain and Education, that is devoted to understanding and improving the ways in which children absorb, retain and apply knowledge.”

“But the innovations have not yet been applied to homework. Mind, Brain and Education methods may seem unfamiliar and even counterintuitive, but they are simple to understand and easy to carry out. And after-school assignments are ripe for the kind of improvements the new science offers.”

Paul offers several examples. As its name implies, “spaced repetition” means exposing students to the same material in shorter sessions over a longer period of time. “Instead of concentrating the study of information in single blocks, as many homework assignments currently do — reading about, say, the Civil War one evening and Reconstruction the next,” she writes, “students are re-exposed to information about the Civil War and Reconstruction throughout the semester.”

Another technique, “retrieval practice,” suggests using tests “not to assess what students know, but to reinforce it.”

“We often conceive of memory as something like a storage tank and a test as a kind of dipstick that measures how much information we’ve put in there. But that’s not actually how the brain works. Every time we pull up a memory, we make it stronger and more lasting, so that testing doesn’t just measure, it changes learning. Simply reading over material to be learned, or even taking notes and making outlines, as many homework assignments require, doesn’t have this effect.”

According to Paul one experiment showed that students using retrieval practice remembered 80% of the vocabulary words they studied, compared to one-third for those using conventional methods. She also describes a technique called “interleaving” which mixes up different kinds of situations or problems to be practiced, instead of grouping them by type. “When students can’t tell in advance what kind of knowledge or problem-solving strategy will be required to answer a question, their brains have to work harder to come up with the solution, and the result is that students learn the material more thoroughly,” observes Paul. Each of these concepts, she concludes, are untapped opportunities to improve student achievement.

Paul’s article is unlikely to be persuasive to anti-homework dead-enders. But she offers a fresh way of considering one of education’s thorniest debates. It’s a must-read for teachers and parents. Paul herself is someone to watch. She is at work on a book titled Brilliant: The New Science of Learning. More importantly she’s proving to be a savvy chronicler of cognitive science and its lessons for educators–of critical importance in a field where “brain-based” is tossed around by hucksters as casually as “new and improved.”


  1. Paul’s article puts a new face–a welcome face, actually–on things teachers have known for a long time. Spaced repetition and retrieval practice are sometimes characterized as “drill and kill.” Basketball coaches and music teachers know, however, that core skills stay sharp with practice, and new ways to understand, review and practice the same old skills really help lock in genuine, useful-in-all-situations proficiency. Paul may be a savvy chronicler of cognitive science, but I’m not seeing a lot of new thinking here. Still, it was a great article. Thanks for sharing.

    What worries me is that in focusing so much on process (which spaced repetition and retrieval practice are) may overshadow the other, equally important, reasons to zero in on the quality of homework: homework should include depth and application of content, assignments that couldn’t be easily completed in school, tasks that use community and family resources or take “think time.” It’s also important for students to understand WHY they’ve been given particular assignments–what’s in it for them, if assignments are completed, besides filling in a box in the grading program?

    You were careful not to use flash words like “project-based learning” or “collaborative problem-solving” or even “critical thinking.” But that’s what Paul is describing when she mentions “interleaving,” which is another take on “scaffolding”–using several previously introduced knowledge strands and skills to answer questions or solve problems.

    Good piece. But Paul is wrong if she thinks good teachers haven’t always considered these principles when assigning homework.

    Comment by Nancy Flanagan — September 12, 2011 @ 11:36 am

  2. OK, so now QUALITY is important? What’s next, MEANING?!?!

    Gee, whiz…

    Comment by Will Fitzhugh — September 12, 2011 @ 11:45 am

  3. Part of the problem is that these strategies, which involve putting individual students on the spot, while effective, are often perceived as punitive and humiliating. There are many ways to avoid this — music teachers and speech therapists know them — and I think pedagogues should concentrate on exploring this aspect more, since real learning can provide a sense of a sense of achievement and competence that can be a source of unparalleled joy.

    Comment by Harold — September 12, 2011 @ 12:09 pm

  4. “When students can’t tell in advance what kind of knowledge or problem-solving strategy will be required to answer a question, their brains have to work harder to come up with the solution, and the result is that students learn the material more thoroughly,” observes Paul.

    My high school students always start out the year asking, “What is the test going to look like?” I think answering that question would be a real disservice. Students really do learn more when they’re absorbing the material instead of learning to take the test.

    I also find it effective to put a few questions the students wouldn’t have studied for on tests. Tests can be learning experiences, too!

    Comment by Alison F. Solove — September 12, 2011 @ 12:44 pm

  5. curious what they think about rote memorization, which my parents had in spades, and my generation a little, but my kids not at all. my parents can still remember poems they memorized in grammar school. i feel like memorization helps with memory in general, and would be an asset. but is not part of the curriculum, or so i’m told every time i ask.

    i often thinks that teachers are hobbled by their principals and curriculum.

    Comment by lola — September 12, 2011 @ 1:50 pm

  6. I would say part of the practical problem here is that what you’d really want to facilitate this is some pretty sophisticated software, but we’re not very good at writing software and have mixed feelings about giving kids computers for school.

    Comment by Tom Hoffman — September 12, 2011 @ 2:25 pm

  7. @Tom Hoffman: the software is already here. Better yet, it’s free and open source. Check out the mnemosyne project, based on the supermemo algorithm. I haven’t seen its use in classrooms reported, but it provides spaced repetition, can serve as a low stakes “learning by testing,” and (if you mix up the various “decks” of flashcards each student has) can provide the “interleaving she writes about.

    Comment by james mink — September 12, 2011 @ 2:44 pm

  8. From the cognitivist Ulrich Neisser:

    “You can get a good deal from rehearsal
    If it just has the proper dispersal.
    You would just be an ass
    To do it en masse:
    Your remembering would turn out much worsal.”

    E. D. Hirsch

    Comment by don hirsch — September 12, 2011 @ 5:18 pm

  9. Nancy is right. These ideas are not new. Nor does one need “brain science” in order to arrive at them.

    Of course a good test can reinforce and even enhance learning. Take a good literature or history or math test, and you find that it challenges you to put together the known material in a new way (or at least a way that has not been introduced in class).

    I don’t mean to disparage brain science, which has a great deal to offer. But I get suspicious when someone presents as an “innovation” (or as a “Mind, Brain, and Education Method”) something that has been known and practiced by teachers for a long time.

    Comment by Diana Senechal — September 12, 2011 @ 5:24 pm

  10. @ Diana @Nancy. Just me perhaps, but I’m cheered when cognitive science provides solid evidence for time-honored practices. I save my skepticism for “brain-based” research that tells us how to turn lead into gold.

    This is no mere exercise in confirmation bias. It is fair to assume that generations of smart teachers arrived at “best practices” not in lab coats but through trial and errors. We tend to seek out the new and novel, when perhaps there is wisdom in the tried and true.

    Comment by Robert Pondiscio — September 12, 2011 @ 5:38 pm

  11. Homework should be an exercise in discipline for adulthood. Most meaningful adult occupations require homework (work to be completed after one leaves the job). Whether you’re a surgeon or a plumber, an attorney or a contractor; if you’re going to be successful at what you do, you’re more than likely going to have work to do at night for the next day. If you haven’t developed the discipline to do a comprehensive job on your adult “homework” you may well be jeopardizing your living.

    One other thought on homework revolves around my orthodoxy; kids are all over the map in school in different subjects. If homework is to be optimized for all the students in the class, it needs to be assigned individually, based on what each student has or has not learned in that subject/class to date. If it isn’t, teachers are feeding into the stupidity of homework as simply busy work.

    Comment by Paul Hoss — September 12, 2011 @ 6:15 pm

  12. Agreed, Robert, but there’s no excuse for calling this an “innovation.” I would have no complaints if she simply said that brain science is explaining the efficacy of some of the oldest and most reliable teaching practices. What would be the harm or shame in recognizing that these practices have been in place for a long time?

    Now, it’s true that many history textbooks could use more “dispersal.” By contrast, it would be hard for a math, language, or music textbook or course to go without it. If you teach a major triad in a music theory course, you can bet your boots you’ll be returning to it many, many times. Same goes for a conjugation in a language class, or a quadratic equation in a math class.

    Comment by Diana Senechal — September 12, 2011 @ 6:26 pm

  13. And that is a delightful limerick from Ulrich Neisser.

    Comment by Diana Senechal — September 12, 2011 @ 6:32 pm

  14. “Spaced repetition” does not have to equal “drill-and-kill”. I firmly believe my student gets far more out of doing 5 tricky Singapore Primary Mathematics word problems than she would out of cranking out 30 equations like I had to growing up. Multum, non multa.

    Comment by Crimson Wife — September 12, 2011 @ 6:54 pm

  15. Wait, how many teachers have ever figured out and employed spaced repetition in any deliberate sense? Spaced repetition isn’t just about reviewing past material on an ad hoc basis, but about reviewing it at increasing intervals over time. See

    Comment by Stuart Buck — September 12, 2011 @ 7:03 pm

  16. I am glad to see the important topic picked up here. I have written a paper for a curriculum conference last year that may be of interest to some. Here is the abstract and below is the link to the complete paper.

    Homework: Its forms and functions revisited

    Homework is seen by many teachers and parents as a useful strategy to promote student learning. However, there are persistent voices claiming that homework is ineffective and should not be used as a teaching and learning strategy. Using data visualisation techniques, the paper introduces a taxonomy of homework. In its examination of current debates about the various forms and functions of homework, this paper draws on the recently completed comprehensive review of the homework literature by the Canadian Council of Learning (2009). A key finding of the Canadian review is that the positive effect of homework on academic performance can be attributed to two interrelated factors: (a) highly developed learning-to-learn skills and (b) intrinsic motivation. In the absence of these personal attributes, homework is unlikely to be successful in fulfilling its desired outcomes. Hence, this paper concludes that investing in the teaching and learning of ‘soft’ skills in formal and informal learning situations is what is most urgently needed.

    Comment by Eva Dobozy — September 12, 2011 @ 8:15 pm

  17. Stuart,

    A great deal of repetition is built into certain subjects. It isn’t ad hoc; it’s inevitable and frequent.

    Yes, there are topics that fade into the background once they’re learned, and teachers often take pains to reintroduce them.

    Wozniak and others have devised systems for repetition of material. These may work very well to certain ends. But the principles have long been in place.

    Now, to learn any subject well, you have to be willing to carry it in your mind, toy with it, recall it even when not prompted. You can’t be in a rush to get it over with.

    Comment by Diana Senechal — September 12, 2011 @ 8:34 pm

  18. If spaced repetition merely meant that some things come up again or that it’s nice to review material, then of course it would be nothing new. But what cognitive scientists are talking about is quantifying the best time(S) to review material to maintain memory. This does not seem to be what teachers were already doing, except in the rarest of accidents.

    From the article I cited:

    “The Bjorks were not the first psychologists to make this distinction, but they and a series of collaborators used a broad range of experimental data to show how these laws of memory wreak havoc on students and teachers. One of the problems is that the amount of storage strength you gain from practice is inversely correlated with the current retrieval strength. In other words, the harder you have to work to get the right answer, the more the answer is sealed in memory. Precisely those things that seem to signal we’re learning well — easy performance on drills, fluency during a lesson, even the subjective feeling that we know something — are misleading when it comes to predicting whether we will remember it in the future. “The most motivated and innovative teachers, to the extent they take current performance as their guide, are going to do the wrong things,” Robert Bjork says. “It’s almost sinister.”

    Comment by Stuart Buck — September 13, 2011 @ 9:34 am

  19. Here it comes…thousands of teachers being docked on their performance evaluations because they failed to provide evidence of “spaced repetition” or “interleaving” in their lesson plans.

    Love the insights that research provides (when it’s not fraudulent or ideologically tainted). Dread the education establishment getting hold of it and enforcing it as law and scripture.

    Comment by James O'Keeffe — September 13, 2011 @ 12:57 pm

  20. I’m with Nancy and Diana on the idea there is nothing new here. Teachers, good ones at least, have always known that the quality of homework is more important than the quantity. That just comes from the simple common sense observation that choosing the wrong problems to assign can result in busy work for students that causes frustration and demoralization, whereas choosing the right problems to assign can result in learning and satisfaction of accomplishment. But I think making good assignments is more an art than a science. And what would be a very good assignment for one student can be something else for another student. And an assignment that brings long term dividends may not seem very satisfying at the moment. Thus conscientious teachers do the best they can. They employ their intelligence, their intuition, their knowledge of their subject matter, and perhaps most of all their experience, to make good assignments. They may not succeed perfectly, but life goes on. Kids do learn, imperfectly, of course, but they do learn.

    A very important question is this: Do the terms such as “retrieval practice” and “interleaving” really add anything to our present knowledge and methods? That is not entirely a rhetorical question. Such terms might be of value, or they might not. But if they are to add something, should they not first be fitted into current practice? To present them as something brand new is not only foolish (and a little insulting), but it disconnects them from present practice and thought.

    Annie Murphy Paul, like almost everyone else, seems supremely confident that the simple term “current practice” is sufficient to refer to what everyone does and what everyone knows. But it is not. And here I fault teachers. There are many, many good teachers. They teach well and effectively, but they do it on an intuitive basis. They cannot tell me how to be a good teacher because they have never verbalized or analyzed to any depth what they do and how they do it. I am as quick as anyone to disparage education professors as lacking in any real body of knowledge of pedagogy. And I am quick to say that the average classroom teacher, with a few years experience, knows more about teaching than an education professor who has been out of the classroom for a few years. But I am also quick to add that if you take an effective and experienced classroom teacher out of the classroom and put him or her in the place of an education professor, he or she would do much the same as the education professor. That is simply because teachers know about teaching intuitively, and only intuitively. They teach intuitively and do a good job. But that should not be enough.

    What is really important, in my mind is that teachers, or at least educational researchers, develop a basis of description and analysis of what actually goes on in real life classrooms and in the heads of real life students and teachers. We need a science of pedagogy, or maybe an art of pedagogy, or just a body of knowledge of pedagogy. Without a basis of description and analysis of real life teaching and learning we’ll continue to think every new term is a new idea or a new understanding, rather than just another candidate for a tired and never ending parade of educational fads.

    I have tried to give some ideas of “current practice” in an article at And I have railed against the sorry state of pedagogy in an article “The Lack of Description In The Study Of Education” at

    Comment by Brian Rude — September 13, 2011 @ 3:14 pm

  21. I agree with Brian that this is more of an art than a science (though I use the word “art” cautiously). I am skeptical about a science of spaced repetition and other such techniques. I read Wozniak’s article and was left unimpressed. (I am curious to know what Dan Willingham thinks of this.)

    Spaced repetition at certain time intervals may indeed be beneficial. But it quickly turns ridiculous if it doesn’t suit the subject or topic. Repetitions should generally make sense within the subject.

    For instance, if one wanted to reintroduce the Civil War during a unit on the Reconstruction, one could look at how people during the Reconstruction looked back on the Civil War. That would give students insight into both eras. There are many other things that can be done. Math teachers arfully bring algebra into geometry, or long division into algebra. Grammar teachers might point out how subject-verb agreement (a concept taught early on) can help the reader navigate a sentence with multiple dependent clauses.

    Now, Wozniak and Paul make an excellent point: at least some of the time, it’s the difficult learning that sticks. I wouldn’t say that’s always true. You need to reach fluency, so you can’t always be struggling. THere should be some ease along the way. But the struggle sure does make its mark.

    Comment by Diana Senechal — September 13, 2011 @ 4:26 pm

  22. The innovation here has yet to be innovated. The innovation would be a kid pulling out his iPad on the bus ride home and spending a half hour (for example) going through a review of spaced repetition of content and skills relevant to his or her studies in all classes, focused on his or her exact progression through the curriculum and personal strengths/weaknesses.

    I don’t believe we’re close to anything like that, or likely to be anytime soon..

    Comment by Tom Hoffman — September 13, 2011 @ 4:39 pm

  23. Thanks, Brian, for your long and thoughtful response. I do agree that much of what good teachers do is intuitive and comes from their own willingness to experiment and pay attention to results. What turned my own practice from all-intuition, all-the-time, was going through National Board Certification, which upped the “scientific” component of analyzing student results and providing credible evidence of important learning, something that’s hard to do with standardized test data.

    I find it interesting that Stuart Buck’s two comments highlight the “almost sinister” failings of teachers whose successes he characterizes as the “rarest of accidents.” Yes, Stuart, teachers *have* figured out that reviewing concepts is best done in increasing intervals and through multiple modalities. They understand that struggling to understand a concept helps embed it more deeply into memory. In fact, that’s one of the principles lying under constructivism.

    As for determining the perfect, research-based intervals for repetition, I am wondering if Buck has ever had to cope with 35 restless students and a mandated curricular pacing chart.

    Comment by Nancy Flanagan — September 14, 2011 @ 11:48 am

  24. The task of reviewing, say, 100 different concepts/facts at just the right intervals for 25 different children (who need review at different times to maximize their learning) seems completely impossible. You’d need to keep track of 2500 different variables for a single classroom, and then somehow differentiate instruction to that micro-level.

    It’s not a slight to teachers to doubt that anyone has ever done this.

    Comment by Stuart Buck — September 14, 2011 @ 1:37 pm

  25. Stuart,

    It is possible and should be done by all. I’ve done it with 32, 2-3 graders.

    Comment by Paul Hoss — September 14, 2011 @ 5:19 pm

  26. You’re kidding, right? You’ve kept track of thousands of different combinations of Student-22-needs-to-review-Concept-157-at-Time-4?

    Comment by Stuart Buck — September 14, 2011 @ 7:50 pm

  27. It has always been my opinion that homework is there to reinforce an idea, rather than present a new one. However, homework too often becomes “Finish this assignment at home since we did not have enough time to finish it in class” for whatever reason. This, then, means that students no longer have the ability to ask questions as they are learning and end up either not completing the assignment or simply doing it incorrectly or ineffectively and have to then be retaught anyway. A more effective use of homework, then, may be to review a quiz students took and find the correct answers, or to read a passage and come up with questions to ask for a discussion the following day. This way, students are not learning new material on their own, but reviewing, researching, or thinking about their learning at home to present to the group the following morning.

    Comment by J. Bisti — September 21, 2011 @ 6:59 pm

RSS feed for comments on this post. TrackBack URL

Leave a comment

While the Core Knowledge Foundation wants to hear from readers of this blog, it reserves the right to not post comments online and to edit them for content and appropriateness.