Homework That Works: Less Volume, More Thinking

Homework is the most universally assigned and least universally examined thing schools do. It gets set because the timetable says homework should be set, marked because parents expect marking, and resented by everyone involved more often than anyone admits. Yet done deliberately, homework is one of the highest-leverage tools a school has. The difference is a handful of principles that most homework policies never touch.

What homework is actually for

Strip away tradition and homework has three legitimate jobs. Practice: consolidating something taught today, because retrieval strengthens memory. Preparation: arriving at tomorrow's lesson ready to use what you know. And independence: learning to manage work without a teacher standing nearby. Anything that doesn't serve one of those three is busywork, and busywork is worse than nothing, because it teaches students that school tasks are performances rather than learning.

The practical test for any homework you're about to set: can you say which of the three jobs it does? "Complete exercises one to twenty" is practice if the exercises were chosen, and filler if they were the next page. The difference is invisible on the surface and enormous in effect.

Less volume, more thinking

The research consensus on homework volume is uncomfortable for the traditionalist: piling on quantity produces steeply diminishing returns, and the returns approach zero (or worse) when students are doing work they cannot attempt successfully. Twenty thoughtful minutes beat two hours of coping. A student who attempts ten problems wrong has practised being wrong ten times, and un-practising that is your problem next lesson.

So cut the volume and raise the thinking per item. Five questions where the third one requires deciding which method applies beats twenty questions of the same type. One "explain the mistake in this worked solution" teaches more than a page of repetition. The planning cost is modest: you're choosing problems, not authoring them, and the best sources for thinking-heavy items are the errors your own class made last week.

The feedback problem, solved structurally

Homework that gets no feedback teaches students it doesn't matter. But fully marking every book every day is the workload trap nobody escapes. The structural fix is to stop treating all homework as written marking. Self-checking homework (answers provided, working required) shifts the first feedback loop to the student's kitchen table. Starter-based feedback turns last night's homework into today's first five minutes: you circulate, you see who's stuck, you reteach on the spot. And auto-graded digital homework handles the objective work entirely, with the class's results visible to you before the lesson begins. Reserve your written marking for the extended pieces where your words actually change something.

Homework that parents can support

In Ghanaian homes, the parental homework question is real: many parents want to help and don't know the modern methods being taught. Two policies close that gap. First, every homework carries a "how to help" line: what the topic is, what a good attempt looks like, and what to do if stuck (try the example, re-read the notes, mark it for the morning rather than guess). Second, keep homework in a predictable place and rhythm, same book, same time expectation, so the household can build the routine around it. Predictability is worth more to completion than any amount of chasing.

The completion problem

Every school has the completion tail: the same few students who never submit. The punitive cycle (punish, resent, avoid) doesn't fix it, because the causes are usually structural: no quiet place to work, no one at home who can enforce, work that's too hard to start. The interventions that work are boringly practical. Make the first question answerable by everyone, so starting is easy. Offer a homework club slot where the work can be done before leaving school. And check completion early (the first week) rather than discovering the pattern at term's end. In LexsEdu, homework set through the Assessments tools makes completion visible per student the moment it's due, and auto-grading returns results the same evening, so the follow-up conversation happens while the work is still fresh in the child's mind.

A homework policy in five lines

If your school's homework policy is a page of generalities, try this instead. One: every homework serves practice, preparation or independence, and you can say which. Two: time expectations are published per level and respected. Three: every homework has a feedback route, self-check, starter review or digital grading. Four: the first question must be startable by everyone. Five: completion is tracked from week one, and the response to non-completion is a conversation before it's a sanction. Five lines, all enforceable, all aimed at learning rather than tradition.

The shift worth making this term

Pick one class. Change one variable: cut the volume in half and raise the thinking per item, or move the feedback from your evening to the lesson starter, or publish the "how to help" line to parents. Watch what happens to completion and to the quality of the work that does come in. Homework earns its place in the school day only when it's doing one of its three real jobs, and the schools that audit honestly find that half of what they assign was never doing any of them. Fix that, and homework becomes what it was always promised to be: the bridge between lessons, walked by the student on their own two feet.

For the assessment thinking that underpins good practice tasks, read formative vs summative assessment, and for the workload arithmetic behind the feedback choices, see cutting teacher workload without cutting corners.

What this looks like across subjects

The principles transfer anywhere. In English, replace "read chapter four and answer the questions" with "find the sentence where the author's attitude changes, and write one line on how you know": a startable, thinking-heavy task with a visible feedback route in the starter. In science, a prediction homework ("tomorrow we'll mix these; write your prediction and why") is preparation in its purest form, and the reveal in class makes the homework feel worth doing. In mathematics, fewer problems with a required method note ("show the step where you decided") beats a page of drills, because the deciding step is the learning. Whatever the subject, the audit question stays the same: which job is this homework doing, and would I stake the lesson on it?