The quid pro quo for small class sizes may not be better attainment, one of the most senior international education analysts told the Scottish Learning Festival last week.
Andreas Schleicher, head of education analysis at the Organisation for Economic Co-operation and Development, contrasted how money was spent in Korea - one of the highest-attaining countries - and Luxembourg, whose attainment is below the OECD average.
The South Korean government had decided it wanted to get the best people into the teaching profession, so it paid teachers very high salaries. In return, teachers had long school days and were expected to perform other duties beyond teaching.
"How does Korea afford it?" he asked a seminar at the SLF. "They have very large classes."
Luxembourg, on the other hand, had small classes, but teachers were paid low salaries. Teachers also had lower teaching hours and did not do things other than teaching.
A country's level of ambition was a key to how it defined standards. To do well, countries should "define what excellence is in education" and make that clear to students, he said, in a review of the OECD's latest findings in its annual publication, Education at a Glance.
"They (the best-performing countries) don't tell teachers what to teach - there is very little prescription for teachers," said Mr Schleicher.
But they provided strong support systems. Norway and Sweden, like top- performing Finland, provided strong support to their teachers, but they did not have strong expectations of excellence, unlike Finland, and focused more on minimum standards. As a result, their performance was not as high as Finland's.
The three most effective forms of continuing professional development were: individual and collaborative research; qualification programmes; and observational visits to other schools - none of which happened very often.
OECD research showed that teachers most wanted CPD which focused on teaching children with special learning needs, how to use ICT effectively, and how to deal with discipline problems.
Grouping pupils according to ability had a negative effect, probably because it took away responsibility from teachers to solve problems, he said.
His organisation had also analysed which initiatives had the biggest impact on raising attainment in science:
- one additional hour of science per week showed a significant effect;
- one hour of homework had less impact than an extra hour of learning in school;
- private tutoring had a negative effect;
- promoting science learning outside school, such as trips to science centres, had quite a bit of impact on performance, even accounting for differences in social background.