Anne Corbett reports on France's attempts to develop a "culture of evaluation". The Lycee Hoche in Versailles is regarded as one of France's most brilliant schools. In 1993 this public-sector establishment put up 351 pupils for the baccalaur at, the multi-subject exam which marks the end of secondary studies and qualifies its holders for university. The success rate for the whole of France in 1993 was 73.3 per cent.
In the literary and natural sciences-led sections, 97 per cent of the students the Lycee Hoche entered were successful. In the maths-led section, first gate on France's royal road to a top career, a full 100 per cent of students passed.
But the Lycee Hoche is also regarded as a school with an easy job. It is a smart school in a smart area, taking the motivated children of successful parents. In France, as in Britain, pressure has been building up to create performance indicators or "value-added" measures that enable schools to be judged not just by their results but by their effectiveness.
In a recently published pioneering work, the French ministry of education has applied three performance indicators to lycees alongside the crude rates of baccalaur at success.
The ministry shows the percentage who pass the baccalaur at in relation to the school's intake in the first of the three years leading to the exam, and the number of baccalaur at options on offer. It also shows the success rate in terms of school-leavers. That is to say, out of every 100 pupils who enter the classe de seconde, does the school let them all stay on, or does it weed out doubtful candidates? And out of every 100 pupils who leave the school how many have been encouraged to stay on an extra year or two if necessary to get the bac certificate? Those are measures of a school's internal practice, and its degree of selectivity.
This French work also attempts to eliminate the incidence of external factors on lycee performance by matching pupils in a particular lycee by age and social class against pupils of the same age and social class on a national basis and on the basis of the academie, the unit of local educational administration.
Reading the Lycee Hoche results in terms of these indicators tells you it does indeed do well. But it is also highly selective.
In France, schools with a similar social intake and age structure will expect 80 per cent of those starting out in seconde to pass the bac. The Versailles academie is only marginally lower at 79 per cent. The Hoche, having weeded out candidates it regards as doubtful, falls below the average, clocking in at 75 per cent.
The results for the Hoche and France's 2,400 other lycees are published in two volumes, one devoted to the general and technological baccalaur at and the other to the vocational bac.
There is also a public demand for league tables; Le Monde de l'Education has been running tables of baccalaur at results for 14 years.
But Claude Thelot, the man behind France's performance indicators, was convinced his country could do better than produce raw tables. He believes it can develop "a culture of evaluation" that will provide useful diagnostic knowledge for teachers, help administrators to introduce change more effectively and give the public a better appreciation of the strengths and weaknesses of schools.
If France has developed value-added measures based on social class, it is because of an "historic commitment" to reducing social and geographical inequalities and a legal obligation to have education policies which favour equality of opportunity.
Thelot, director of the ministry of education's evaluation and forecasting section, is adamant about the code to be respected in evaluating school performance. It must be useful to those who run the school as well as those who want to use it.
"An evaluation simply designed for outsiders degenerates into league tables, " he says. It should also be able to indicate an "added value" measure alongside exam results.
The ministry is now providing all the public lycees, and all the private lycees under contract, with a set of standard indicators and a technique for each school to calculate its own performance, which heads can use as management material.
These lycee indicators are the latest stage in work which has been going on intensively since 1990, when Thelot was appointed, to upgrade the ministry's statistical base so that it becomes an analytical tool.
Other notable examples of the new-style evaluation designed to help teachers and policy-makers and to inform the public are the mass testing of pupils in French and in maths in the third year of primary school, the first year of lower secondary school and the first year of upper secondary schooling; and the annual publication of the 30 indicators of the education system's performance in terms of its cost, its activities and school performance.
Thelot says, optimistically: "Once evaluation becomes a habit and state of mind, on top of the classroom benefits, the public debate will be less sterile."