The First Minister's views on internal selection within schools are worrying.
Not only do they appear out of sync with what his previous education minister was promoting through A Curriculum for Excellence, but they also appear to breathe life into one of education's most enduring myths: namely, that pupils can be separated into classes on the basis of some test of prior attainment and that predictions of what they can be expected to achieve can be accurately made - on the basis of no evidence whatever.
Almost a decade ago, the inspectorate made a similar attempt to promote setting in the face of research evidence. It published a report in 1996, Achievement for All, but omitted any references to the literature review it had commissioned to inform the report. The review, by two respected academics, Wynne Harlen and Heather Malcolm, appeared a year later, and it was no surprise to find that their conclusion did not support the inspectorate's advice that setting would improve attainment.
There was some limited evidence that some higher-attaining pupils did better in top sets in maths but that lower-attaining pupils did better in mixed ability classes.
However, there was also evidence that the negative impact on learners in bottom sets of low aspirations, low expectations, low self-esteem and a sense of being less valued, would be likely to impact on attainment.
The most prolific researchers in this field in recent years have been Susan Hallam and Judith Ireson of London University. They have carried out a number of studies on the impact of setting on attainment, pupil motivation and self-image. Their 2006 study found that pupils' attitudes to setting were affected by their set placement, type of school, socio-economic status and gender.
One of their studies, based on data from 6,000 Year 9 (S3) students in 45 mixed secondary comprehensives, showed no significant effects of setting in English, mathematics or science. In addition, the effects on higher and lower attaining students were not consistent in the three subjects.
Socially disadvantaged students achieved significantly lower grades, and girls achieved higher grades than boys, especially in English.
What this evidence is telling us is that setting will not provide a simple organisational solution to a complex problem.
In the recent past, intelligence tests, which sought to predict people's future success, were rightly discredited. They discriminated against people on grounds of race, gender and social class. Setting in our secondary schools results in these same groups being over-represented in lower sets.
Setting is a form of internal selection. Perhaps we should realise that "ability" is not the only factor in pupils' success. Application, motivation, resilience, "stickability", self-theories and self-belief all play a part. Putting people together on the basis of test scores and assuming that they form a homogeneous group will not provide the answer to the problem of underachievement.
Setting should no longer be the default position in Scottish schools.
Brian Boyd professor of education at Strathclyde University, was a member of the review group which produced A Curriculum for Excellence