Most of my detentions at school were for a lack of homework. I remember going through my excuses for non-completion on the bus to school. Today, if I were to ask my teenage self his opinions on homework, like many other children, the reply would be an insistence that it was a waste of time; a complaint that the chore was unnecessary; and a demand for its abolishment.

How many teachers share such a view? The relevance of homework to learning and pupil attainment, and its impact, continues to be a source of controversy among pupils, teachers and parents alike. Moreover, research investigating the benefits of homework on pupil attainment shows no concrete agreement on either its relevance or efficacy. Although many studies show a positive correlation between time spent doing homework and pupil attainment, other evidence suggests that homework may have negative effects on learning.

So what should schools do? The Herts and Bucks Challenge Partners hub, which is a group of schools in Hertfordshire and Buckinghamshire working to inform best practice and school improvement, set up a joint research project examining the impact of homework on learning during the 2017-18 academic year. The aim of the project was to research the existing literature, meet to discuss this work and consider the effect on pupils’ attainment of the homework we were setting.

Importantly, the project emphasised the need to understand homework in our own contexts, particularly working towards a greater understanding of how the type of homework activities we set affect learning.

Identifying the best types

Despite protests from my inner teenager, we concluded relatively quickly that homework was “generally” beneficial. We knew this from Harris Cooper et al (1989; 2006) and their gargantuan meta-analyses of previous studies on homework; from the Department for Education’s own research (Sammons et al, 2014) and from various other studies, such as Cowan and Hallam (1999).

We were also aware that most of the research on the benefits of homework did not question the impact that different types of homework activity have on pupil attainment - but there is, thankfully, one study that does do this. Research by Darling-Hammond and Ifill-Lynch (2006) found that teachers should make homework purposeful and relevant, which suggests that some types of homework activity will be better than others.

As a result, we wanted to unearth which types of homework activity were best for learning, especially as this would help us to decide which types of homework to set regularly for our pupils.

We found that researchers on homework tended to identify at least three or four types of homework activity (for examples, see Rosário et al, 2015; Dobozy, 2010; and Darn, 2007). These types included:

• Practice homework, which involves questions, exercises or tasks on content already taught.

• Preparation homework, which involves preparing for upcoming lessons.

• Traditional revision, which includes rereading information from lessons.

• Extended homework, which includes longer homework activities, such as project work.

The working group also suggested two other types:

• Creative homework, which could include building motte-and-bailey castles for history, for example.

• Finishing-off homework, which is the straightforward completion of unfinished work.

These were included because it became clear that we set these types of homework regularly, especially at key stage 3.

To ascertain which of these activities had the most impact, we decided to canvass the views of our colleagues and pupils, as well as monitor how students were responding to the different homework activities set in two control groups. Furthermore, my colleagues met regularly to compare notes and see if we could reach a consensus on which type of homework activity worked best.

Finding a consensus

By and large, we all seemed to agree with the research that practice homework had the greatest effect on the development of cognitive, or rather academic, attainment. For example:

• Studies by Cepeda et al (2008), Carpenter (2014) and Brown, Roediger and McDaniel (2014), among others, empirically demonstrate the importance of improving knowledge retrieval via practice-based questions.

• Sixty-nine per cent of teachers at our school, The Reach Free School (TRFS) in Rickmansworth, Hertfordshire, who were involved in the homework working group believed that - after setting different types of homework in their subject areas - practice homework had the biggest impact on pupils’ attainment. (This was followed by preparation homework, which was mostly advocated by English and creative-arts teachers.) Creative and finishing-off homework activities were seen to have the least impact on attainment.

• In a survey of pupils at TRFS, 74 per cent said they felt practice-homework activities had improved their assessment grades the most. Although a subjective opinion, the pupils were asked to complete different types of homework and then qualitatively assess which ones most improved their knowledge, confidence and exam skills. They did, however, use their assessed work to make this judgement.

• In the two trial groups, 91 per cent and 92 per cent of pupils respectively were on target over the “practice only” trial period, as opposed to 45 per cent and 79 per cent respectively prior to the trial period, when homework activities were mixed.

• The Herts and Bucks Challenge Partners hub research group largely concurred with my school’s findings. But it is worth noting that Joe Carey, head of Year 8 at Sir John Lawes School in Harpenden, suggested that its staff and pupils favoured homework that fostered freedom and independence, as opposed to more straightforward practice or preparation homework assignments.

Practising recall is best

This research, albeit limited and carried out by “teacher-researchers”, as opposed to academics, suggests that some types of homework activity are ineffective and that more activities should be focused on practising knowledge recall, knowledge application and exam-style questions on content that has already been taught.

Still, these findings are largely focused on teacher and pupil perceptions, and I would acknowledge that more research needs to be done to demonstrate an empirical effect on pupil attainment. Also, I fully accept that some types of homework activity will benefit pupils in different ways at different phases, in different subjects and in different educational settings. As such, the conclusion drawn here should be seen as a generalisation.

But it is worth noting that, as a somewhat reluctant learner myself who regularly boycotted homework as a teenager, I left secondary school with only two GCSEs. It was only when I repeated everything at a further education college that I realised the relevance of working away from the classroom. I hope my current pupils avoid my mistake and do their homework.

Andrew Jones is assistant headteacher at the Reach Free School in Rickmansworth, Hertfordshire. He ran the Herts and Bucks Challenge Partners hub along with Mike Garvey, assistant head of St Clement Danes School in Chorleywood