Just before Christmas, I attended an event to mark the retirement of a highly-regarded London headteacher. He had joined his school in the 1970s on a lower rung of the management ladder. His description of how the school population had changed over the years was fascinating. It was a salutary reminder that we should be cautious about interpreting changes in school performance without knowing whether the “performers” have been changing too. It was also a reminder of the huge demands placed on some heads and staff to meet the fast-changing needs of pupils and local communities.

The school was formed by an amalgamation of a secondary modern and a technical school. Its pupils in the mid-Seventies were predominantly white working-class boys, with Greek Cypriots the largest minority group. Whilst still in the process of development as a comprehensive school, it became co-educational and had to adapt its curriculum, ethos and facilities to accommodate girls. While doing this, it began in the Eighties to see a rapid rise in the number of Bangladeshi children, many arriving direct from the Indian sub-continent, and had to develop provision to meet their needs.

While doing this, it began to see an ever more diverse intake from all over the world.

Today, the school community continues to mutate as it becomes increasingly popular after primary transfer while still welcoming newcomers later on from a wide range of countries. Children come and go as families, often in temporary accommodation, and often move on. Some parents come to London to work or study for a limited period, then return overseas.

The experience of this particular school is certainly not unusual in London and similar stories could also be told outside the capital. And although the preceding account gives some impression of the scale of change such schools must cope with, it does not adequately describe the complexities of the challenge - especially today.

Consider this: some children from overseas have been highly-educated in their country of origin and are fluent in English; some are highly educated but speak little English; some have had little formal education; some had their schooling severely disrupted by civil war or lengthy and tough journeys to the UK.

Of course, all schools are aware of cohort variation - the way in which the characteristics and ability of children can differ significantly between year groups - but in many urban schools, with complex, fast-changing intakes, the variations will be especially dramatic and certainly make it impossible to draw simplistic conclusions from changes in exam results across the years. Changes in the nature of pupils can lead to improvement as well as decline in a school’s exam results but neither outcome in itself shows whether the school is doing a better or worse job.

Change in a school’s population can happen quickly where a school with many vacancies begins to admit large numbers of new pupils in all years. When this occurs, children taking examinations at 16 will have been in the school for different lengths of time and this may be reflected in the results.

At one school with which we are currently working, 48 per cent of pupils who had been attending since the beginning of Year 7 achieved 5 Cs or better at GCSE last year - close to the national average. However, the school’s performance was affected by those who arrived in subsequent years: only 28 per cent of those late arrivals got five A*-C GCSEs. Many schools have data which replicate this picture. Not only do aggregate results conceal such variations, but they also conceal good progress made by late arrivals with low achievement. Some children who arrive late with little English perform poorly at first but advance rapidly and achieve spectacular success later on.

I am doubtful whether the Government’s value-added tables can ever take proper account of the complexities of change in school communities. And what about the value added by teachers to children who come and go before examinations even take place.

However, to give ministers their due, work on pupil mobility started by Charles Clarke (in his first incarnation as education minister) and carried forward by Estelle Morris has led to a nationwide initiative by the Department for Education and Skills to identify effective ways of managing high levels of movement.

In the meantime, I am waiting to see how many of these rapidly changing schools will be deemed worthy of “earned autonomy”. This will apparently bestow on schools “new freedom to shape the curriculum” in a way that reflects the special needs of its pupils and local community.

Interestingly, the head I mentioned earlier thought, with good reason, that he had been trying to do that for the past 30 years.

Dr Janet Dobson is carrying out research, funded by the Nuffield Foundation, on pupil mobility in secondary schools and would be interested to receive observations from schools on the above issues. Address: Migration Research Unit, UCL, 26 Bedford Way, London WC1H 0AP. Email: janet.dobson@ucl.ac.uk