Does Homework Actually Work? What Cooper's Meta-Analyses Found — and Why the Grade Level Matters

Lead

"Did you do your homework?" is the default evening prompt in many households. Homework has been treated, largely without examination, as something that improves children's learning. Schools believe it; families accept it.

Researchers, however, have spent more than 40 years producing a careful answer to that question. Their conclusion diverges somewhat from common sense: the effect of homework on academic achievement varies enormously by grade level.

This article steps back from "how to do homework better" and asks the more fundamental question: does it work at all?

Background: Three Types of Homework

Researchers have long distinguished three purposes for homework:

Practice: reinforcing and consolidating classroom content through repetition
Preparation: previewing upcoming material before class
Extension: exploratory tasks that go beyond the textbook — research projects, creative work

These three types place different cognitive demands on children and work through different mechanisms. Treating all homework as a single category, and asking whether "homework" generically improves achievement, is already a simplification. Yet most research has done exactly that — in an attempt to address the broadest possible question.

Cooper's Meta-Analyses: The Core Finding

Educational psychologist Harris Cooper published a foundational in 1989 and an updated version in 2006, covering research from 1987 to 2003 [1,2]. Together, these analyses have served as the primary reference point for homework research.

The 1989 version synthesized 120 studies. The 2006 update added further evidence. The core finding is that the relationship between homework and achievement differs dramatically by grade level:

Elementary school students: correlation between homework quantity and achievement is approximately zero (r ≈ 0) [1,2]
Middle school students: small but detectable positive correlation (r ≈ 0.07) [1,2]
High school students: moderate positive correlation (r ≈ 0.25) [1,2]

A meaningful positive association between homework and achievement appears only from high school onward. For elementary school children, the data show essentially no academic benefit from homework quantity.

Cooper's explanation: elementary school children are still developing the self-regulation skills and time management capacities that make homework an effective format for independent learning. Repetitive tasks done while tired may also produce confusion rather than consolidation [1].

Cooper and colleagues derived a practical guideline from this finding — the "10-minute rule": a weekly guideline of approximately 10 minutes of homework per grade level (roughly 10 minutes in first grade, 30 minutes in third grade) [4]. This is not a recommendation to assign no homework; it is a recommendation to take the question of quantity seriously rather than defaulting to more.

The Parental Involvement Paradox

A meta-analysis by Patall, Cooper, and Robinson (2008) synthesized 14 studies on parental homework involvement and produced a counterintuitive pattern [3]:

Monitoring and correcting involvement ("Did you do it?" "That answer's wrong." "Here, do it this way"): negatively associated with academic achievement
Autonomy-supportive involvement ("When do you want to start?" "I'm nearby if you get stuck"): positively associated with academic achievement

Why does monitoring backfire? Researchers point to two mechanisms. The first is that over-involvement by a parent can undermine a child's sense of — the feeling that "I can do this myself." The second is more direct: when a parent supplies the answer, the practice that homework is supposed to provide doesn't happen.

Evidence also shows that high-quality homework assignments are associated with improved achievement, while increased quantity or inappropriate parental intervention erodes the effect [5]. Both the design of the homework and the type of parental involvement shape outcomes — not just the hours spent.

What Homework Can and Cannot Change

Even if the direct effect on test scores is limited for younger children, proponents of homework often argue for non-academic benefits: building work habits, time management, responsibility, and the foundations of self-regulated learning.

Trautwein et al. (2006) found that homework engagement itself can predict learning attitudes [6] — so the habit dimension is not trivial. The caveat is that if building good habits is the goal, the design of the homework matters. Tasks that are entirely joyless and poorly calibrated to the child's level may cultivate the opposite: an aversion to learning.

Practical Takeaways

Know the volume guideline. The "grade times 10 minutes" rule is a useful reference point. If the homework being assigned substantially exceeds that, raising the question with the teacher is a reasonable step.

The "nearby but not intervening" stance. Neither hovering nor disappearing. Something like "let me know if you get stuck" and then being available is close to what Patall et al.'s "autonomy-supportive involvement" describes in practice.

Ask about process, not completion. "What was hard today?" points attention toward the inside of the learning. "Did you finish all of it?" points toward the output. Both have a place, but the first tends to open a more useful conversation.

Summary

Forty years of research does not support "more homework is always better," nor does it say "homework is worthless." For younger elementary school children, direct effects on academic achievement are not reliably confirmed, and there is no evidence that more is better. Monitoring and corrective parental involvement tends to backfire; autonomy-supportive involvement tends to help.

The value of homework lies less in its immediate effect on grades than in the habit of engaging with learning on one's own — which means that how it is approached, at home and at school, shapes its real effects more than the quantity assigned.

References

Cooper H. Synthesis of research on homework. Educ Leadersh. 1989;47(3):85–91.
Cooper H, Robinson JC, Patall EA. Does homework improve academic achievement? A synthesis of research, 1987–2003. Rev Educ Res. 2006;76(1):1–62. doi:10.3102/00346543076001001
Patall EA, Cooper H, Robinson JC. Parent involvement in homework: a research synthesis. Rev Educ Res. 2008;78(4):1039–1101. doi:10.3102/0034654308325185
Cooper H. The Battle Over Homework: Common Ground for Administrators, Teachers, and Parents. 2nd ed. Thousand Oaks: Corwin Press; 2001.
Dettmers S, Trautwein U, Lüdtke O, Kunter M, Baumert J. Homework works if homework quality is high: using multilevel modeling to predict the development of achievement in mathematics. J Educ Psychol. 2010;102(2):467–482. doi:10.1037/a0018453
Trautwein U, Lüdtke O, Schnyder I, Niggli A. Predicting homework effort: support for a domain-specific, multilevel homework model. J Educ Psychol. 2006;98(2):438–456. doi:10.1037/0022-0663.98.2.438
Fernández-Alonso R, Suárez-Álvarez J, Muñiz J. Adolescents' homework performance in mathematics and science: personal factors and teaching practices. J Educ Psychol. 2015;107(4):1075–1085. doi:10.1037/edu0000032