Primary pupils may be quite right to demonise David Blunkett, the Education and Employment Secretary, for insisting they do homework.

While it is of proven value at secondary level, its effect at primary level is less clear, according to a review of research presented to a conference of the British Psychological Society in Exeter last week. It can create extreme tensions in families, breeding resentment or dependency in the child.

But the Department for Education and Employment, which has just issued draft guidelines on homework, said modest amounts of homework were “essential” to the drive to raise standards in literacy and numeracy.

The report, by Susan Hallam and Richard Cowan of London University’s Institute of Education, draws its findings from a review of more than 100 studies, mainly American, because there is little British research.

It says much of the research on the effectiveness of homework is inconclusive. There is some evidence that homework which prepares for future work or revises previous work is of more benefit than just finishing off working started in class. There is little evidence that homework develops responsibility in pupils.

Mr Blunkett’s guidelines say four and five-year-olds should do 10 minutes of reading and 10 minutes of other homework each day, rising to 30 minutes at the age of six and 40 at the age of 8.

“Is Homework Important for Increasing Educational Attainment?” Dr Susan Hallam and Richard Cowan