Homework Research: What Studies Show About When It Helps and When It Hurts

Homework is among the most universal features of American education and one of the most debated. Parents argue about whether their children have too much or too little; researchers debate whether homework improves learning or merely consumes time; and educators navigate pressure from both directions. Research on homework effects has produced findings that are more nuanced than either strong advocates or strong critics typically acknowledge, with grade level, subject matter, and homework quality emerging as critical moderating factors. Harris Cooper, whose meta-analyses of homework research have been the most comprehensive and frequently cited, finds that the relationship between homework and academic achievement is positive and statistically significant at the high school level, smaller and less consistent at the middle school level, and essentially nonexistent at the elementary school level. This grade-level pattern has been replicated in other analyses and is one of the most consistent findings in homework research. The explanation typically offered is that older students have developed the organizational skills and self-regulation necessary to benefit from independent practice, while younger students have not. Subject area matters as well. Homework effects are stronger and more consistent for subjects involving procedural skills that benefit from repeated practice, including mathematics and foreign language, than for subjects requiring conceptual understanding or creative work. Research on mathematics homework finds that practice of previously learned skills produces achievement gains, while homework involving novel material that students have not yet mastered in class produces less consistent benefits and may produce frustration and misconceptions. The amount of homework is a critical variable that research suggests is not simply a matter of more being better. Research on the relationship between homework quantity and achievement finds diminishing returns at high volumes and, at extreme levels, negative associations with wellbeing and school engagement. Studies of homework duration find that beyond approximately one to two hours per night for high school students, additional homework is not associated with additional learning and is associated with increased stress, sleep disruption, and reduced family time. Research on students in high-homework environments documents significant time pressures and health consequences that have led some districts to implement homework quantity limits. Quality of homework matters more than quantity. Research on homework design finds that assignments that require students to practice and retrieve previously learned material, that are closely aligned with classroom instruction, and that students can complete with reasonable success are more effective than assignments that are merely long, complex, or disconnected from classroom learning. The testing effect, meaning the learning benefits of retrieving information from memory compared to passive review, suggests that practice retrieval homework is particularly valuable for consolidating learning. Equity implications of homework are significant and underappreciated. Research finds that the conditions under which students can complete homework vary enormously by family socioeconomic status. Students from low-income families are more likely to lack quiet study space, reliable internet access for online homework components, and parents with time and resources to support homework completion. When homework assumes these resources, it may disadvantage students from lower-income families and amplify rather than reduce existing achievement gaps. Schools that assign similar homework to students with very different home circumstances may inadvertently compound inequity. Research on homework and family dynamics finds mixed effects on family relationships. Parents who help with homework report both positive experiences of connection with their children's learning and significant stress about homework demands. Studies of parental homework involvement find that high-pressure involvement that does the work for children is associated with poorer outcomes than supportive involvement that provides guidance while maintaining children's agency. Family homework involvement also varies by socioeconomic status and parent education level, with more educated parents better positioned to support academic homework in ways that promote learning. This creates another mechanism through which homework can reproduce rather than reduce educational inequality. The research suggests a homework policy that emphasizes purposeful practice assignments aligned with classroom instruction, limits homework volume especially in elementary school, provides flexibility for students with limited home resources, and uses homework primarily for retrieval practice and application of already-learned skills rather than as a venue for first exposure to new material. The debate about homework is unlikely to be fully resolved, but the research provides clearer guidance than either abolitionists or maximalists typically acknowledge.