For generations, homework has been a non-negotiable part of education. It’s the sacred ritual of school life; the proof of learning, the test of discipline, the gateway to success. Or is it? As we deepen our understanding of neuroscience, learning psychology and the evolving role of artificial intelligence (AI), the question arises: Is our traditional approach to homework truly serving students, or is it time to rethink the model?
The Homework Myth: More Work, More Learning?
One of the most enduring beliefs in education is that more practice equals better mastery. The logic seems sound —students reinforce what they learn in class by repeating it at home. However, research suggests that this assumption is flawed.
“It’s the sacred ritual of school life; the proof of learning, the test of discipline, the gateway to success. Or is it?”
John Hattie’s Visible Learning meta-analysis (2017) found that homework has a modest overall effect on achievement (d = 0.29), with negligible benefits in primary school and moderate impact in secondary school. In younger students, homework’s impact is virtually zero, largely because they lack the executive function skills to work independently in meaningful ways.
The prefrontal cortex; the brain’s centre for planning, reasoning and self-regulation; does not fully develop until the mid-20s (Diamond & Lee, 2011).
“Expecting young learners to self-direct their learning without scaffolding is like handing them a car and hoping they figure out how to drive.”
What’s more, cognitive science research (Roediger & Butler, 2011) highlights that retrieval practice with immediate corrective feedback is far more effective than repetitive homework drills. If students practise incorrectly, they are reinforcing errors, not mastery. Without expert coaching, neural pathways form, making them harder to unlearn later.
Stress, Sleep and the Science of Overload
In high-pressure schooling environments, homework has become synonymous with stress. The OECD’s Education at a Glance (2014) found that students in countries with excessive homework loads report higher stress levels and lower well-being. The stress hormone cortisol, when sustained, disrupts the connectivity between the prefrontal cortex and other brain regions, impairing attention, problem-solving and memory (Diamond, 2011). In short: more stress, less learning.
Equally troubling is the effect of homework on sleep. Research by the National Sleep Foundation (2019) confirms that adolescents require 8–10 hours of sleep per night, yet students with heavy homework loads often get significantly less. Sleep is not just ‘rest’; it is when the brain consolidates learning and strengthens neural connections. A tired brain is a brain that forgets.
The Role of AI: Cheating or Opportunity?
With the rise of AI, homework as we know it is facing an existential crisis. If AI can complete a student’s homework better and faster than they can, does it still serve a purpose? Some argue that AI-driven homework leads to laziness and dishonesty. But an alternative perspective is that AI exposes the inefficiency of traditional homework models.
Rather than banning AI, forward-thinking educators are integrating it into learning. AI should be a thinking partner, not a shortcut. By structuring assignments that require students to critically engage with AI—fact-checking, analysing and debating outputs—homework can become more relevant, insightful and aligned with real-world skills.
So, Should We Abolish Homework?
Not necessarily. What the research tells us is that homework must evolve. A more effective approach prioritises:
- Deliberate practice over busywork – Short, targeted tasks that reinforce key concepts.
- Scaffolded assignments – Providing structured guidance to develop independent learning skills.
- Time-conscious learning – Recognising that learning requires rest and reflection, not just repetition.
- AI-integrated tasks – Encouraging students to engage with AI critically rather than avoiding it.
- Retrieval-based learning – Using techniques like spaced repetition and active recall to improve retention.
Reimagining Homework for a New Era
Homework, in its traditional sense, is a relic of an outdated model. The most effective learning happens not when students are overloaded but when they are challenged in the right ways. This means rethinking not just how much homework we assign, but what kind of work students are doing and whether it genuinely enhances learning or merely serves as a compliance exercise.
Education is not about clinging to tradition for tradition’s sake; it’s about evolving in ways that best serve students. If AI has exposed the cracks in the system, let’s take the opportunity to build something better.
“The question isn’t whether we should keep homework or throw it out entirely. The real question is: How do we make it work for the world we live in today?“