I’m sure at some point this is a question we’ve all pondered, especially when chasing that elusive piece of maths homework from “litte Dafydd” in year 6.
A 2017 meta-study* of nearly 370,000 learners from KS2, KS3 and KS4 across 10 countries has found that for science and mathematics, the picture is (unsurprisingly) quite complex.
In terms of effect size (d) (remembering that d=1 is equivalent to 1 years additional teaching or 2 GCSE grades and d=0.5 is equivalent to ½ a year’s additional teaching and 1 GCSE grade), where Hattie considers d=0.4 as the “hinge point” above which an intervention is “interesting” for a school to consider, the team found average effect sizes of:
Key Stage (Effect size)
What does this imply?
Parental involvement (0)
Frequency of homework (0.29)
Marking for effort (0.41)
Time of homework (0.46)
Marking for completion (0.61)
Marking for correctness (1.2)
Overall the study points quite strongly towards the impact on outcomes being limited to KS4 age learners and with an effect size of 0.58 this corresponds to an increase in 1 GCSE grade. At KS2 and KS3 there is little to link homework in general to an increase of educational outcomes. The authors further subdivide the impact of homework by the style of intervention. Parental involvement and increasing the frequency of homework have very marginal impact on standards. Interventions focused on assessing the completion of a challenging homework task and marking for overall correctness / grades have the most impact. With an effect size of d=1.2 this corresponds to two additional grades at GCSE or an additional years teaching.
Of course homework carries a whole range of potential benefits that are not aligned to measurable outcomes; confidence, resilience, pride, rehearsal and enrichment and these need to be considered alongside the impact on attainment. However, one thing is clear, at KS4 where the impact of homework on attainment is the greatest, marking for correctness offer the biggest impact of any style of homework intervention.
*The full paper can be obtained here (paywalled) http://www.sciencedirect.com/science/article/pii/S1747938X16300628