The UK should be wary of following the North American example of setting a national aptitude test for university admissions, a senior official at the Organisation for Economic Co-operation and Development warned this week, writes Paul Hill.
Barry McGaw, the OECD's director of education, told a conference in Cambridge there were doubts about the effectiveness of the SAT, which is used by universities in the US and Canada.
He said there was evidence of "coaching" for the test in schools and of results reflecting students' social and ethnic backgrounds. "Introducing an aptitude test is no measure of what pupils are meant to be learning," he said.
His comments follow a government decision this summer to run a pilot scheme of a UK SAT involving 50,000 A-level students.
Calls to introduce a SAT to help spot candidates with "hidden talent" - bright pupils from poor-performing schools - followed last year's Schwartz report.
Professor McGaw told the conference organised by Cambridge Assessment, formerly the University of Cambridge Local Examinations Syndicate, that research evidence showed that the greatest improvement in performance came from tests to identify what teaching a pupil needed at school.
Onora O'Neill, principal of Newnham College, Cambridge, warned that the current school exams system was blighted by "bogus" statistics and assumptions, including the idea that all A-level subjects are equally hard.
Measuring school performance by the number of A to C grades had created a "perverse incentive" for schools to encourage pupils into subjects where "A-level points are more easily come by", she said, citing the fall in the number of students sitting German and maths as evidence.
Baroness O'Neill warned that exams carried "too great a weight". Results were used to measure pupils' achievement, school and teacher performance, as well as being used as a selecting tool by universities and employers.