When I started in 1990 we already had a computer system but ours was designed to integrate transport planning, so every child had a unique pupil number. That was very unusual then. It gave us the opportunity to use the data to raise standards. We started value-added in 1993, using reading test scores to measure improvement in national curriculum tests and GCSE. That’s now fairly commonplace. The Government is about to produce value-added tables for secondaries. Importantly, we started it in primary schools. We gave data to schools in graphical form, so they didn’t have to be statisticians.

Teachers really wanted to know how much progress children were making. Schools with disadvantaged catchments discovered that they were adding more value than schools in the leafy suburb - it was a real boost to teachers’ morale and confidence.

We’ve now gone on to things like benchmarking: comparing how you are doing with how you should be doing. It’s an incentive for schools to strive for something greater. We provide the data, but the self-improvement must come from them.

In 1997, we started professional dialogue with schools. Link advisers sat down with heads and chairs of governors to review performance, given the school’s cohorts. They set priorities for improvement. Then numeric targets were set. It was a powerful tool. In 1999, when the Government introduced education development plans, they encapsulated what we’d been doing.

We’ve had our own software package, Sparks, to help schools to analyse data. It cuts down hugely on the amount of work schools have to do.

When national curriculum tests were introduced, we looked at the OMR sheets and thought there must be a better way, so we gave teachers a computer disk that would do everything for them.

It’s all grown very quickly since then because the leading schools are always coming up with ideas to help us improve the software, so we are always three years ahead of most of our schools. There’s a danger, though, if you get too far ahead.

I think we’re moving into a more national system now. We’ve got hugely powerful computers in schools so they don’t need us to analyse information for them. There’ll be a lot more direct transmission of data between schools and government which is realising that you can’t wring much more out of certain testing techniques. You can see this in the key stage 3 strategy with the emphasis on thinking skills and on teaching for learning.