Children who got the most maths homework help from parents scored lower, but improved the most over a year
Studies disagree on whether parents helping with maths homework helps or hurts. This Norwegian study looked at both how well children did and how much they improved. Children whose parents helped more often scored lower in maths, but over one year they made the biggest gains. The likely story is that parents help more when a child is struggling, so "help goes with low scores" does not mean help causes low scores.
| Paper | Parents' homework help versus children's achievement level and development in mathematics |
| Authors | Arve Fiskerstrand, Markku S. Hannula |
| Published | 28 April 2025 · Research in Mathematics Education (Taylor & Francis), online 28 April 2025 |
| Study type | Longitudinal correlational study (structural equation model, two waves a year apart) · Peer-reviewed |
| What we read | Full text read. We read the full open-access article (the University of Helsinki repository copy). |
| Summarized |
What they did
The authors reused anonymised data from the Norwegian SPEED project. Children took the same kind of 40-question multiple-choice maths test twice, a year apart, and children and parents answered surveys about homework help, parents' education, and how much parents valued and encouraged achievement. The analysis used a structural equation model with a latent change score for maths development.
Who took part
796 children aged 10 to 15 and their parents in Norway who took part in both data collections and answered all the relevant questions (out of 3,015 children who took part in the first data collection of the original project).
What they found
- More parental maths homework help went with a lower maths achievement level (r = −.51).
- But more help went with more improvement from the first year to the second (r = .32).
- Parents' education (r = .31) and their achievement attitudes (r = .25) went with higher achievement levels, but slightly with less improvement (r = −.14 and −.19).
- Children with above-average help went from 58.58 to 68.53 points (a 17.0% gain, Cohen's d = .567); children with below-average help went from 68.89 to 76.05 (10.4%, d = .433).
Effect size: Correlations from the structural model: help and achievement level r = −.51; help and achievement development r = .32. Within-group gains: d = .433 to .567.
Limitations
- Correlational: it cannot show that help caused the gains.
- Children who started lower had more room to improve, which may account for part of the bigger gains.
- Only 796 of 3,015 children had complete data, and parents who answer surveys may differ from those who do not.
- Homework help was measured as how often it happened, not what kind of help it was.
- One country, ages 10 to 15; the authors caution about generalising internationally.
What this means for learners and parents
- If you help a lot because your child is finding maths hard, a low score is not a sign that your help is failing.
- Look at progress over months, not just the level on one test.
- The study did not measure what good help looks like. Other research suggests help that leaves the thinking with the child works better than doing it for them.
Put it into practice (free tools)
- How to help with homework without giving the answer
- Maths homework you don't understand yourself
- Times tables out loud (free tool)
Citation
Arve Fiskerstrand, Markku S. Hannula (2025). Parents' homework help versus children's achievement level and development in mathematics. Research in Mathematics Education (Taylor & Francis), online 28 April 2025. https://doi.org/10.1080/14794802.2025.2493841
Related summaries
- Children felt they got more homework help and less freedom than their parents thought, and autonomy support went with better results · Mia M. Maurer et al., 2026 · Longitudinal correlational study