Homework Policy: What Research Says and What Actually Works

Homework is one of the most debated topics in education — and also one of the most researched. The gap between what research says and what most schools actually do is significant. Here's what the evidence shows and how to design a homework policy that aligns with it.

What the Research Actually Says

The most comprehensive review of homework research is Harris Cooper's meta-analysis, which has been updated multiple times over three decades. The key findings:

• Elementary school: little to no relationship between homework and academic achievement. The evidence for homework in grades K–6 is weak.
• Middle school: modest positive correlation with achievement when homework is assigned appropriately.
• High school: stronger positive correlation, particularly for students who complete homework.

The headline: homework benefits high school students more than younger students, and it only benefits students who actually do it.

The "10-minute rule" (10 minutes of homework per grade level per night) comes from Cooper's research as a reasonable upper bound. A third grader's homework should take about 30 minutes. A tenth grader's might take 100 minutes. Most teachers assign significantly more than this.

What Undermines Homework Effectiveness

Homework fails to produce learning when:

• It's too easy: practice problems on fully mastered skills don't build anything. They just fill time.
• It's too hard: if students can't do the work independently, they either give up, copy, or have a parent do it for them.
• It's not checked: assigned homework that's never collected or discussed communicates that it doesn't matter.
• Volume exceeds capacity: students with 4+ hours of homework disengage and the work quality collapses.
• It's penalized inequitably: students without study space, quiet, or family support are penalized for conditions outside their control.

What Makes Homework Effective

Homework works when it is:

• Practice on skills already introduced: homework is consolidation, not introduction. Never assign practice on something not yet taught.
• Short and focused: 20–30 minutes of meaningful practice beats 90 minutes of mixed, unfocused tasks.
• Completed by the student (not a parent or tutor): this seems obvious, but assignment design affects who actually does the work. Tasks requiring students to observe something, interview someone, or solve problems using their own understanding are harder to outsource.
• Reviewed and connected to class: homework that's collected, checked, and referenced in the next class becomes part of the learning cycle.
• Differentiated when possible: struggling students need different practice than advanced students.

Designing a Homework Policy That Works

For elementary teachers: keep it minimal. Reading (20 minutes per night) has robust research support at all grade levels. Math fact practice has some support. Project-based homework sent home on weekends is often more parent work than student work. Be skeptical of elaborate at-home projects.

For middle school teachers: 30–45 minutes total (not per class) is appropriate. Coordinate with colleagues if possible — four teachers each assigning 30 minutes creates a 2-hour burden. Skill practice, reading assignments, and written responses are reasonable. Test prep packets rarely produce the gains teachers hope for.

For high school teachers: nightly reading and problem practice are appropriate. Be realistic about load. Students in multiple AP courses face unsustainable homework burdens that produce anxiety, sleep deprivation, and learned helplessness — none of which support learning.

The Equity Problem

Homework policies that assign grades for completion are inequitable by design. A student who spends the evening at a parent's second job, helping younger siblings, or managing a chaotic home environment cannot do four hours of homework every night. When homework completion affects grades, you're measuring socioeconomic conditions as much as academic performance.

Solutions:

• Separate academic performance (mastery grades) from homework completion (work habits grade, tracked separately)
• Design homework for independence: tasks that require only what was taught, not outside resources or adult support
• Build in class time for homework questions — students who had no help at home need a chance to get it

The Bottom Line

For K–6: consider whether the homework you're assigning has a plausible mechanism for improving achievement. If it doesn't, reconsider whether to assign it. Twenty minutes of recreational reading has stronger evidence than most homework assignments.

For 7–12: homework can support learning, but only when it's the right type, the right length, and genuinely completed by students. Audit what you're assigning against the research before adding more.

The goal of homework is not to signal rigor. It's to extend and consolidate learning that began in class.