What Does the Research Actually Say About Homework?

Homework is one of education's most contested topics. On one side: parents and teachers who believe practice outside school is essential for learning and academic development. On the other: researchers and reformers who argue that homework — especially excessive homework — produces stress, family conflict, and little academic benefit. Both sides are drawing on real evidence, partially.

The honest picture is more nuanced than either camp typically acknowledges, and the nuance matters for how teachers design homework assignments.

What the Research Actually Shows

The most rigorous synthesis of homework research (Harris Cooper's meta-analyses, among others) finds a consistent pattern by grade level:

High school: Moderate positive relationship between homework and academic achievement. Students who do homework generally perform better on tests and grades than students who don't.

Middle school: Weaker positive relationship. Some benefit, but the effect size is smaller than at the high school level.

Elementary school: Little to no relationship between homework and academic achievement. The research consistently finds that homework assigned to elementary-age students produces minimal academic benefit compared to what that time would produce through other activities — play, family time, reading for pleasure.

This doesn't mean all homework is bad. It means homework's value depends heavily on grade level, assignment design, and what it replaces.

When Homework Helps and When It Doesn't

The research distinguishes between types of homework:

Homework that works: Practice of skills already taught and understood, applied to new examples. Reading for sustained engagement and comprehension. Long-term projects with meaningful work occurring in small chunks over time. Pre-reading or background building that activates prior knowledge for the next day's lesson.

Homework that doesn't work: Homework assigned primarily for accountability (to show students did something) rather than learning. Homework that introduces new concepts without instruction — students working on material they haven't learned produces error practice rather than skill building. Busywork that requires time without producing thinking.

Homework that harms: Homework that's consistently too much, producing chronic sleep deprivation, especially in adolescents whose sleep needs are high and whose natural sleep rhythms conflict with early school start times. Homework that is inaccessible to students because of limited home resources, language barriers, or family circumstances that don't allow for quiet study time.

The Equity Problem

One of the most underappreciated issues in the homework debate is equity. Homework assumes conditions at home that many students don't have: a quiet space, adequate lighting, internet access, a parent who is available and able to help, and time not consumed by work, care responsibilities, or economic precarity.

When homework completion is measured and counted, it systematically disadvantages students in difficult home circumstances. A student who doesn't complete homework because they share a room with four siblings and spent two hours babysitting after school isn't less motivated or less capable — they lack the conditions that homework assumes.

This doesn't mean homework should be eliminated, but it does mean teachers should be clear-eyed about what homework compliance actually measures — is it learning capacity, or is it home circumstance?

Designing Homework That Actually Helps

If you're going to assign homework, design it deliberately:

Match it to grade level: Elementary homework should be minimal — fifteen minutes, primarily independent reading. Middle school homework can build to thirty to forty-five minutes. High school homework should be purposeful, but the widely cited ten-minute-per-grade guideline (ten minutes for first grade, twenty for second, etc.) is a reasonable benchmark even if not universal.

Only assign practice on things students already understand: Students who don't understand the concept can't practice it correctly — they practice their misunderstanding instead. Homework is for consolidating and applying learned skills, not for learning new ones.

Build in choice and relevance where possible: Homework that connects to students' lives or interests (reading a book they chose, applying math to something they care about, writing about something that matters to them) produces more engagement than decontextualized practice problems.

Design for accessibility: Can this homework be completed without internet access? Without a parent's help? Without specific materials students may not have at home? If not, reconsider the design or provide the materials.

Be thoughtful about grading it: Grading completion of homework that students may have completed with parents' help, copied from peers, or done under duress doesn't reflect learning. If homework is practice, assess its learning outcomes on in-class tasks rather than grading the homework itself.

Your Next Step

Audit the last week's homework you assigned: what was the purpose of each assignment? Did it require practice on things students understood? Was it accessible to all students regardless of home circumstances? Would the learning outcomes be better assessed by an in-class task? Most teachers find that this audit reveals one or two assignments that were busywork — and a few that were genuinely purposeful. Start by eliminating the busywork. The remaining homework will be better for students and easier to manage for you.