Actually, writing is never easy: in case anyone doesn’t know, it’s the hardest work around.” (Truman Capote.) I do not understand why we are in such a national spin about children’s writing. The apparent problem is that far fewer pupils reach the target level 4 in writing in key stage 2 than they do in reading.

The real problem is the target: it assumes, quite falsely, that all subjects, and all their different aspects, are equally tough. Yet most adults find writing harder, too. And a one-size target looks pretty dotty at the chalk-face.

But we can - and should - expect to improve children’s writing. I have just finished surveying the writing achievements of 17 wonderful schools across England; they often had gritty chalk-faces.

None of my sample had privileged intakes, yet the average gap between their level 4-plus achievements in reading and writing was only nine percentage points.

One school had no gap, and no pupils below level 4. And an inner city school where 87 per cent of the pupils were speakers of English as an additional language had a gap of exactly nine points, an achievement many more leafy schools might envy.

They must have been getting something right. What they were doing was complex of course, a blend of leadership, a passion for language and literature, tight organisation, attractive resources, and some quite excellent teaching.

But a number of messages stand out clearly. In particular, they had adopted the National Literacy Strategy wholeheartedly, but on their own terms, and they did much more than it requires.

Their own terms included being flexible about how writing is taught. In particular, it meant finding more time for teaching and practising writing, than a sequence of five identically structured literacy hours can conceivably provide.

It also meant being bold about the levels: the sequence in which they are taught, and the priorities that seem to be implied by the way they are listed in the strategy framework (word-sentence-text).

My evidence suggests that schools that are highly effective with writing give the highest priority to text-level work.

They develop children’s writing by inviting them to write: often, at length (extended writing), and for purposes that make sense to their pupils. In their teaching, word and sentence-level activities serve the purposes of known text-level tasks. The purposeful writing task determines the word and sentence level study. Where writing is most effective, word and sentence study are not permitted to become discrete or dominant, and are never undertaken in their own right, nor for their own sake. These are lessons that will be needed once again at key stage 3, where lesson “starters”, (ie, discrete word and sentence study) are being actively promoted by the strategy this term.

These schools also excelled at doing more than the strategy requires, but I can only list two further themes here. Perhaps the most significant lay in the time and attention that they also gave to reading. As literacy guru Frank Smith once put it, “writing needs reading”; the survey schools were unusual in the extent to which reading fed their written work.

In particular, they werecreatively dissatisfied with the diet of extracts that a regime of unvaried literacy hours is apt, inadvertently, to promote.

They read complete texts, and whole novels in particular. They found, too, that, when it came to writing, traditional tales from different cultures helped powerfully with story structure.

And they gave a special emphasis to the close assessment of children’s progress in writing. When they had assessed, they took action, and intervened early, especially with underachieving boys.

Graham Frater is anindependent educationadviser and former HMI.His survey of KS2 writing will be published by the Basic Skills Agency later this month:Tel: 0870 600 2400