The six-year project began in 1996 with reception-age children in six schools and followed them for two years. At the end of the first year, the results were encouraging - children who used the early reading framework had a mean reading age seven months higher than in six comparison schools, and six months higher than the mean chronological age.

More than 120 Essex primary schools have now been involved in early-reading research, says Sue Kerfoot, the authority’s project manager, and news of its success is spreading.

The focus is on helping children break words into constituent sounds (segmentation), and put sounds together to make words (synthesis). This they do in short, regular bursts, sounding out words as a whole class - with the teacher first modelling the exercise for them - then repeating it together and then by themselves.

A key part of the project - and here it differs markedly from the literacy hour - is that the exercises must be done three-times-a-day, for 10 to 15 minutes, to minimise the amount of information that children forget. The project schools have reached a compromise with the national strategy - one session is accommodated within a slightly abridged literacy hour; the others take place later in the day.

Hearing children read - once a day if possible - is another strand of the project and, as in the literacy strategy, children are taught to read the most common 100 words by sight.

Kathy Dowsett, headteacher of Highwood community primary school in Colchester, where the project is in its second year, says she had some reservations about giving children “an unremitting diet of phonics”, but is impressed by what has been achieved.

“Early-years and reception children have learnt to read core words very quickly,” she says. “As a spin-off, teachers have noticed that these children are more independent about approaching writing, because they have that secure phonic base.”

Diana Hinds