DONALD Hirsch wrote recently (Platform, January 3) on an OECD project which claimed to provide “conclusive”, “unequivocal” research evidence that selective secondary schooling and setting by ability in comprehensive schools are ineffective. They apparently fail to improve pupil performance and are socially regressive.

Few may now wish to go back to a selective, two-tier system but there is a hardening view that inferior performance by comprehensive schools has to be tackled head-on and the Scottish Executive has made moves in this direction. There has been encouragement for experiments in setting in core subjects and it has been suggested, not for the first time, that this would benefit all levels of ability.

Research suggesting that setting is, after all, not a viable policy has therefore to be scrutinised closely. An OECD project of this magnitude would normally command respect but, on this occasion the research appears to have been seriously flawed. The project tested the reading skills of 15-year-olds across Japan, Hungary, Germany, France and Finland, the UK and other countries. The researchers were unfazed by comparing such widely differing systems and cultures with so many variables, even though it was unlikely they would be comparing like with like.

Literacy was selected for the tests presumably as a skill that reflects achievement across other subjects. This is, however, a questionable assumption in post-primary education. The researchers relate schooling to parents’ socio-economic status but ignore pupil behaviour and discipline, parental support and national cultures of high or low achievement in schools.

Most disturbingly, they assert that old-style selective, two-tier schooling and setting in the comprehensive system are comparable. This is simply untrue. Moreover no distinction is made between selective state schools and fee-paying schools exclusive to a well-heeled parental group. Selective schools in general tend to be well-staffed and well-disciplined but fee-paying schools enjoy better equipment, smaller class sizes and (in boarding schools) a total learning and living environment that day pupils necessarily adapt to. A comprehensive, state system - with or without setting - can in no way be compared to the independent sector.

With all this, the researchers assume that any move to set pupils by ability abolishes the difference between non-selective and selective systems, and between state and independent schools. Not true.

Worse-performing comprehensives have a much higher staff turnover, poor buildings and equipment, less well qualified staff and varying levels of supply teaching. They can have serious behavioural and motivational problems, many associated with an unsupportive home environment or lack of space at home for serious study. They can generate a culture of low achievement, while independent schools encourage high achievement by pupils in the upper range of ability.

Disturbingly, the researchers fail to identify meaningful categories of pupil or to take account of other major factors that contribute to low performance. Research in the social sciences is vitally linked to controlling variables, clearly identifying groups to be compared and comparing like with like. It also has to avoid preconceived ideas on deprivation, which smack of social engineering rather than rigorous research.

In Scotland, there are no grounds for confusing setting in core subjects in a comprehensive with across the board streaming or especially with a two-tier selection system. Yet the argument that setting is ineffective and regressive is based on the assumption that, by setting pupils, a comprehensive changes its skin and identifies with independent schooling.

For a long time it was indeed assumed that non-selective entry meant pupils would not be set by ability in any subject. Mixed-ability teaching became the norm. The argument was that what the ablest gained through setting was paid for by the less able. But evidence has since accumulated that mixed-ability teaching sacrifices both the ablest and least able.

he claim that “setting widens the gap on inclusion” is blinkered, perverse and unproven. Setting has simply nothing to do with “selective education”.

Moreover, countries that differ in their views on selection also have widely differing patterns of pupil behaviour, parent support and employment needs. Far from being socially restrictive, if setting in core subjects works, abler pupils could quit the cycle of social deprivation and less able pupils from all sections of society would emerge with useable skills as they receive more appropriate teaching.

If we are to regenerate comprehensive schooling, all the factors have to be looked into: indiscipline and poor behaviour, parent support, stresses that force good teachers to leave, motivation. If setting by ability has the slightest chance of breaking the mould, it has to be considered and it should not be rejected out of sheer prejudice or on the say-so of poorly conducted research.

The Executive should mount a national programme to identify, pilot and implement best practice. It should tackle head-on the problem of our academies of failure and underachievement. It should tackle the worst of them and reduce their class sizes; inject gifted teachers to vary the mix in overstressed staffrooms; introduce parent-pupil school contracts; teach pupils in ways more appropriate to their abilities.

The last thing Scotland needs is botched research giving an air of respectability to underperforming schools.

Samuel Taylor is retired professor of French at St Andrews University and was heavily involved in language developments in schools.