Using homework as a tool to reinforce what students are taught in school is a convention as old as the classroom itself. However, what homework sessions should look like is less of a settled topic. 

A new study published March 16 by the American Psychological Association’s Journal of Educational Psychology analyzed the math and science homework habits of more than 7,000 Spanish students to determine the frequency, time and effort dedicated to homework as well as how often students required help to complete the work. 

The researchers concluded that how homework is done is a much more critical factor in academic performance than how much homework is assigned.

“One of the most important results we observed with this data is that the relationship between homework and academic performance is not linear,” said Javier Suárez-Álvarez, co-author of the paper and a Ph.D. student at The University of Oviedo in Spain. “Sometimes the most effective is not the most efficient.”

According to Suárez-Álvarez, students who spent 60–70 minutes on mathematics and science homework experienced the most benefit. Although more time spent did result in better performance, the returns diminished exponentially after 90 minutes. However, the researchers found that it was not uncommon for some students to be assigned closer to two hours of homework per night, almost double the minimum effective dose. 

“Of course your results will improve a little bit by spending more time on homework, but we observed that the benefit you’re going to obtain is so small,” Suárez-Álvarez said. “The optimal time is around an hour.”

Suárez-Álvarez said the difficulty level involved and whether or not students require help tackling the assignments is a much more important factor in academic success. The tendency was that students who worked independently with minimal help from parents or teachers to complete their assignments scored in a higher percentile on standardized tests.

“Research in the psychological and scientific literature about self-regulated learning establishes that the main key to developing self-regulated learning is to work through problems on your own,” Suárez-Álvarez said. “If we can get teachers to assign homework regularly and systematically, we can give students the opportunity to work autonomously and develop the skills to manage and handle the homework—the problems, the situations—by themselves.”

This means the homework being assigned must rest somewhere between being challenging enough to keep students engaged and motivated to complete it, yet not too difficult that they are stumped and cannot reach the answers without consistently asking for help. According to Andy Isaacs, a senior research associate in the physical sciences division and director of the Center for Elementary Mathematics and Science Education at the University of Chicago, this is a challenging balance to achieve.

“This is, I would say, the dominant paradigm for how you teach mathematics: You take a problem, show the student how to solve it and then practice a bunch of problems using the same method,” Isaacs said. “This is essentially a very passive thing. The kids aren’t really autonomous because they’re taking the method that the teacher showed them and they’re imitating it. What we’ve found in our work is that one huge shortcoming of this method is in real life, if you get a problem, it isn’t preceded by a teacher showing you how to solve it.”

People tend to think of mathematics and science education as eternal and unchanging, but researchers are constantly learning new things about how people learn, Isaacs said. Textbooks and curricula need to incorporate the latest findings in order to be the most effective, and determining the proper “how” elements of homework assignments such as amount and frequency is not only a constantly moving target, but also one that requires specialization for students of different ages and learning abilities.

Diane Briars, president of the National Council of Teachers of Mathematics, said having research that contradicts commonly held beliefs about math education is sometimes met with resistance from an educational system built on long-held learning standards. 

“This study talks about how it’s not necessary to assign huge quantities of homework, but it’s important that assignments are systematic and regular,” Briars said. “We’ve known this. This has been the result of research for a long time. People strongly believe that if I want to learn something, I have to give you a lot of practice all at once. It’s going to make more sense if you do 500 problems better than only 10 spread out over time.”

However, there is data that suggests working in a collaborative setting does hold benefits for learning and retaining information from homework, Briars said. Having the opportunity to analyze how others arrive at answers  and the process they go through to solve problems can help increase understanding of the material. 

“Homework is part of the broader picture of how we improve instruction,” Briars said. “Students need to be engaged in making sense of the mathematics. They need to be solving problems, not listening to somebody lecture. They need to be building procedures out of conceptual understanding so it’s not only showing the steps, it’s understanding why the steps work.”