News

The Problem With Homework: Not Much Evidence on Whether It Works

By Matt Barnum | September 1, 2016

It is a rite of passage for students, parents and teachers alike. Students must dutifully do (or refuse to do) their nightly allotment of homework, complaining about it either way; parents must dutifully nag their kids to finish their homework; and teachers must dutifully assign and grade it.
But not everyone thinks this ritual is unavoidable, including one second-grade Texas teacher, Brandy Young, who recently wrote in a note to parents: “After much research this summer, I am trying something new … Research has been unable to prove that homework improves student performance. Rather, I ask that you spend your evenings doing things that are proven to correlate with student success. Eat dinner as a family, read together, play outside, and get your children to bed early."
Young’s note to the 21 parents in her class quickly went viral, tapping into an apparently pent-up frustration, with many feeling that homework is at worst a waste of time and at best an unnecessary and stress-inducing burden. Not that homework doesn’t have its fans: A recent poll found that most parents feel the amount of homework that their children get is about right, with some even wanting more.
Young elaborated on her thoughts in a piece for the Huffington Post, saying she didn’t set out to create the great homework debate of 2016 but was glad that she did.
So what does research say on the topic of homework?
Shockingly little, as a matter of fact. As I began to dig into this topic, I quickly discovered that there’s a dearth of good evidence out there that could uncover the effects of homework.
Take one oft-cited 2006 overview of past research on homework’s impact. It concludes that there is “consistent evidence for a positive influence of homework on achievement” and that “a stronger correlation existed in Grades 7–12 than in K–6.”
This would seem to align with common sense: Homework has a positive effect, particularly in later grades when the work gets more complex.
Hold up, though. Note the word “correlation”: Just because kids who do, or are assigned, more homework also have better test scores, that doesn’t tell us whether it’s the homework causing the higher achievement. Maybe homework is useless, but high-performing students are more likely to do it anyhow. Or perhaps teachers are more likely to give extra work to kids who are struggling.
The research review acknowledges this complication: “All studies, regardless of type, had design flaws.”
Harris Cooper, co-author of the research review and a Duke professor, said he generally thinks there is significant value in homework, but he agreed there were problems with previous studies. He points out that his research found six studies that could plausibly isolate the impact of homework, with all but one showing consistently positive results.
A half-dozen studies may sound solid, but there are red flags here too. All of them are quite old — ranging from 1985 to 1997 — and, oddly, none was published in an academic journal. That might be because they all appear to have significant flaws in how they were set up and carried out, which Harris and his co-authors document.
“Methodological considerations make it difficult to draw strong causal inferences from [the studies’] results,” they write. “Still the results are encouraging because of the consistency of findings.”
In other words, the studies aren’t great, but at least they all agree.
The bottom line is that these studies are small in number, quite old at this point and methodologically dubious, so it’s hard to put too much stock in them.
So what about newer research on homework? Surely with the large number of educational researchers, many have tackled the impacts of homework.
Not so much, as it turns out. There are just a handful of good pieces of evidence on the subject.
One careful study from 2011 compares how the same eighth-grade students performed when assigned to a teacher who gave a lot of homework versus one who gave less. The research found that the kids with more homework did much better in math but no better in English, science or history. Although the study was done somewhat recently, it uses national data from 1988.
The same large positive effect in math also shows up in a 2007 study, particularly for struggling students, and suggests that past research underestimated the value of homework. Notably, it looked only at math and used the same 1988 data set as the previous study.
A 2012 paper used three different methods to examine the relationship between homework and how eighth-graders fared on international math and science tests. All three approaches suggest that more homework produces higher scores.
Probably the most rigorous research of all randomly assigned college students homework in introductory economics classes at several public colleges in North Carolina. The effects of both being given and actually doing homework were positive across several dimensions: grades, exam scores, class retention. But since this 2013 study involved college students, its applicability to K–12 education is unclear.
A few other studies use fancy statistical approaches to try to isolate the impacts of homework. They all find positive effects for math homework in upper grades, though once again, the data used is quite old.
And that’s about it in terms of rigorous evidence for or against homework.
Cooper tells me that he hasn’t seen any studies that carefully examine potential negative impacts of homework, like stress or family disputes. One survey of wealthy high school students found that those who reported doing more homework were more stressed out and had worse health, but again, the study couldn’t show that more homework was to blame.
Tom Loveless, of the Brookings Institution, agrees that there’s been remarkably little research on homework that can identify its causal impact. “There are very few serious researchers who have been interested in this topic, and I don’t know why this is,” he said. Loveless says the paucity of evidence means district leaders and policymakers should be wary of setting any sort of mandate on the amount of homework assigned to students.
That leaves teachers in a tough spot. While there’s some support for math homework in later years, in other grades and subjects, there’s precious little evidence one way or another to go on.
Almost every research paper in any subject calls for more research, even in areas that have been extensively studied. Here, however, it’s really needed; researchers ought to devote a lot more time to answering this question.
Until then, the homework rite of passage will continue — at least in classrooms other than Brandy Young’s.
Submit a Letter to the Editor