What Makes Homework Worth Assigning (And What Doesn't)

Homework is one of those teaching practices that gets defended by tradition more than by evidence. "Students need to practice at home" is true in narrow conditions and false in many others. Before assigning work to follow students out the door, it helps to know exactly what makes homework worth the time it takes from students' evenings.

The research on homework is messier than most teachers were trained to believe. John Hattie's meta-analyses found homework in elementary school has near-zero effect on achievement. In secondary school it has a moderate effect — but the type of homework matters enormously. The benefit comes almost entirely from practice on skills that are already understood, not from first exposure to new material or from projects that require resources students may not have at home.

That narrows what homework is actually good for.

Homework That Works

The clearest case for homework is deliberate practice of a skill a student has already been taught and mostly understands. The goal is fluency, not introduction. Students who have learned long division in class benefit from practicing it that evening. Students who are still confused about long division will only reinforce their confusion by practicing it wrong at home.

This distinction — practice vs. introduction — is one teachers sometimes blur when they assign homework to cover material they didn't get to in class. That puts students in the position of teaching themselves, which works for self-motivated students with good resources and fails everyone else.

A second legitimate use is low-stakes retrieval practice. Short written recalls ("write three things you remember from today's lesson without looking at your notes"), brief quiz-yourself exercises, or quick application of a formula. This isn't just busy work — spaced retrieval is one of the most robust findings in cognitive psychology. The next-day bump in retention is real.

A third use is reading that genuinely can't happen in class because it would take too long. Longer texts, especially in humanities courses, require time that 50 minutes doesn't provide. But the follow-up matters: if students can't come in the next day and discuss the reading, the reading was just an obligation, not a learning event.

Homework That Doesn't Work

Introduction to new material. If you're assigning homework that requires students to watch a video or read a chapter before you've taught the concept, you're assuming they can do independently what the classroom environment is there to support. Flipped classrooms can work, but only when the introductory exposure is well-designed and students come back with questions you address before they practice.

Projects that require expensive resources. Diorama assignments, poster projects, printed research reports — these are equity problems disguised as assignments. A student without a printer, craft supplies, or a parent who can drive to the store after 8pm is being assessed partly on resources, not learning.

Work that can't be meaningfully checked. If you assign 40 problems but only have time to check five, students learn that 35 problems don't matter. If the homework isn't being used — to inform instruction, to diagnose misconceptions, to give feedback — it isn't teaching anything except that school assigns things it doesn't care about.

Long readings with no scaffolding for struggling readers. Independent reading at home requires independent reading skill. If your class includes students who are still developing fluency, sending them home with dense text produces avoidance, not learning.

Designing Homework Deliberately

The most useful question before assigning homework isn't "what should I assign?" but "what will I do with this tomorrow?" If the answer is "nothing, I just want them to practice," ask whether that practice could happen in class. If the answer is "we'll discuss it," make sure the discussion is worth the work. If the answer is "I'll review it and adjust my next lesson," then the homework is actually integrated into your teaching.

Feedback on homework is where most of the learning happens, and most teachers don't have time to provide it. The workaround isn't skipping homework — it's designing homework that builds in self-checking, peer checking, or class-wide debrief that makes the individual work visible.

The Equity Problem You Can't Ignore

Homework assumes students have a quiet place to work, stable lighting, family support or absence-of-disruption, and enough cognitive bandwidth after school. For a significant portion of students — those with jobs, caregiving responsibilities, unstable housing, or families in crisis — those conditions don't exist.

This doesn't mean assigning no homework. It means designing homework that can be done in 20 minutes on a phone if necessary, that doesn't require outside resources, and that can be made up without catastrophic grade penalty when it doesn't happen.

The student who didn't do the homework isn't always the student who didn't care. Often they're the student carrying the most weight outside your classroom.

A Simple Standard

Before you assign anything to take home: Can it be completed in 20 minutes or less? Does it require skills already taught in class? Can students complete it without resources beyond a pencil and their notes? Will you do something meaningful with it tomorrow?

If all four answers are yes, assign it. If not, redesign it or don't assign it at all.

The goal is learning, not the practice of doing homework. Those are related, but they're not the same thing.