The Homework on Homework: Part One

As I promised earlier, I am reading the studies cited by Robert Marzano Classroom Instruction that Works. The first study I picked up is “The Effects of Homework on Learning: A Quantitative Synthesis” from Journal of Educational Research, November/December 1984 (full citation at end of this post).

Alfie Kohn’s claim in his rebuttal of my post at The Faculty Room was that none of the studies cited by Marzano, et. al. in the chapter “Homework and Practice” showed that “homework was beneficial for students.” Kohn accuses Marzano of misrepresenting the research on homework.

The focus of the Paschal, Weinstein, Walberg, 1984, which is not one of the five studies Kohn mentions in his criticism of Marzano, is a synthesis of “empirical studies of homework and of various homework strategies on the academic achievement and attitude of elementary and secondary students.” In the abstract of the study, Paschal, et. al. state: “About 85% of the effect sizes favored the homework groups. The mean effect size is .36 (probability less than .0001). Homework that was graded or contained teachers’ comments produced stronger effects (.80).”

As I said, this meta-analysis does not appear to be one of the five studies Kohn mentions in his post when questioning Marzano’s research, but it is in a chart on p. 61 of Marzano entitled “Research Results for Homework.” Perhaps it is not considered by Kohn because it is a meta-analysis or synthesis rather than original research itself. I do think it has interesting things to say about the effects of homework, however. Paschal, et. al. note that “[e]xtensive classroom research on ‘time on task’ and international comparisons of year-round time for study suggest that additional homework might promote U.S. students’ achievement.” However, the authors also note that writing on the subject of homework has largely characterized homework as “unwholesome, professionally unsupervised, or allow[ing] the children to practice mistakes.” Paschal, et. al. acknowledge that attitudes regarding homework seem to change depending upon a variety of factors.

Paschal, et. al. examined “15 studies that compared students with various qualities and amounts of assigned homework. These included the most frequent comparison, of students who were assigned and those who were not assigned homework.”

The authors conclude that “[t]he corpus of evidence shows a moderately large average effect size [0.80] of assigned homework that is commented upon or graded.” However, the authors also acknowledge that “much of the voluminous, 70-year-old literature on homework is opinionated and polemical, and surprisingly few methodologically adequate studies have been conducted.” I can attest to the veracity of the first part of that statement, given my own recent experience. Kohn was the only respondent to my original post who even brought up research.

Obviously, I want to read the five studies cited by Kohn and Marzano, as those studies seem to be at the heart of the contention between the two, but I felt this meta-analysis made it fairly clear that some homework was better than no homework. Marzano’s conversion table on p. 160 translates a 0.36 effect size to a percentile gain of 14. In other words, the average student who does homework (at least, according to my interpretation of the meta-analysis) will have score 14 percentage points higher on a standard bell curve measuring student achievement.  An average student who does homework that is graded and receives feedback on that homework will have a score over 28 percentage points higher on a standard bell curve measuring student achievement.  Sounds good to me.  I encourage you to read the study yourself and draw your own conclusions, too.  Feel free to leave them in the comments.  I will be reading other studies and sharing my conclusions here, so if you are interested in the great homework debate, check back.

Marzano, Robert J., Debra J. Pickering, and Jane E. Pollock. Classroom Instruction that Works. Alexandria, VA: ASCD, 2001.

Paschal, Rosanne A., Thomas Weinstein, and Herbert J. Walberg. “The Effects of Homework on Learning: A Quantitative Analysis.” Journal of Educational Research. 78.2 (1984): 97-104. Professional Development Collection. EBSCO. Weber School Library, Atlanta, GA. 7 March 2008. <>.

Related posts:

7 thoughts on “The Homework on Homework: Part One

  1. Two of the things that frustrate me in the whole homework discussion/debate are:

    1. What counts as homework? There doesn't seem to be much rigor in the definition.

    2. Does the age of the student affect the results? In other words, is homework equally effective in first grade and tenth grade?

  2. Liz, we have the same frustrations. As far as I can tell:

    1. You're right. Folks don't have an idea what homework is. I thought it was one thing and quickly realized that a lot of people didn't (apparently) agree that what I was defining as homework was actually homework. To me, homework is work not done in class. Period.

    2. The answer to this one is yes, but folks who oppose homework don't like to discuss it for some reason. My thinking is that "yes, but…" makes them feel as if their argument has been weakened. But even Alfie Kohn acknowledged some homework in upper grades is OK, although I still think he is too limiting with regards to the kind. He mentions, for example, that we can or maybe even should read novels in class rather than as homework.

    The study I read focused on grades 4-10, unfortunately, so I can't speak about early grades yet. It is my belief that homework is necessary in high school, and Grant Wiggins agreed with me on that one. Coming from a high school English background, he said, "I cannot imagine an education that does not require students to read and write outside of class." However, elementary school, particularly lower grades, is probably a different story.

  3. I agree that homework is necessary in high school (and to a lesser extent in middle school). You can't get good enough in math without problem sets; you can't get good enough in a 2nd language w/out outside practice; in English & History you just don't have enough time in class to use it reading, and so on.

    The difficulty I see is that teachers in the discrete disciplines often don't coordinate — so an individual child might be overwhelmed by duelling deadlines.

  4. Too true! Which is something Grant also suggests — that schools work together to research and devise a policy for their school. Students have accused teachers at my school of "ganging up" to punish them with a lot to do, but I have told them, quite candidly, that the truth is that we didn't know they already had assignment X, or we would have planned differently. One thing I have tried this year in implementing UbD is more long-term assignments. It's due two weeks from today — pace yourselves. That kind of thing. I think it's working for students.

  5. At the risk of wading (again) into a conflict in which I'm not sure where I fall. You might consider that the sorts of students who do their homework are also the same sorts of students who do better at school. It's not the homework making them better, it's the fact that they are more responsible making them do the homework. Post hoc and all that.

    Still, I should say I give homework and would have to see something pretty impressive before I give it up.

  6. Nate, you might want to check out the actual study itself. It looks at results by the variable of achievement level (low, middle, mixed, and not reported). However, I never had the opportunity to take statistics, and I wasn't 100% sure what to make of the results. I'll see if I can get the help of a statistics teacher on Monday.

  7. wow

    I had no idea there was research comparing graded to ungraded homework (and I think I have Marzano's book).

    This is a major issue in my own district where teachers stop "collecting and correcting" homework around 3rd grade.


Comments are closed.