The biggest-ever UK research project into teaching and learning will challenge the Westminster government's obsession with academic performance - and also the increasing use of "thinking skills" in schools.
Professor Andrew Pollard, director of the economic and social research council's pound;32 million teaching and learning research programme (TLRP), said it was pushing the social aspects of learning very hard.
He told last week's pedagogy conference in Cardiff: "The big picture is all to do with learners as people, and education and its purpose. It is this that is going to challenge the obsession with performance to provide a more balanced view of teaching and learning."
He said people's attitudes towards learning are shaped by their experiences. But this developmental model has so far played second fiddle to the target-oriented economic model of education policy favoured in Westminster.
"We are arguing we need both of these, and policy has been driven too much by the economic model," he said. He added: "History is moving out of a phase dominated by the obsession with performance, targets and structures.
We have got a more balanced, authentic and evidence-informed understanding of learning that is emerging. From that comes a more sensible model of teaching."
Professor Pollard, based at London university's Institute of Education, said findings emerging from the TLRP were also proving critical of teaching methods based on thinking about multiple intelligences and pupils' learning styles.
Many schools are promoting different teaching methods to accommodate pupils' supposed learning styles - for example, kinesthetic or visual - or promoting thinking skills using brain skills and mindmapping.
But these methods can imply pupils have fixed learning needs or styles - "rather than temporary capacities that can be built on and changed", he said.
He was backed up by David Stone, head of Corpus Christi RC high school in Cardiff, who has been seconded to the Assembly government to work on its school leadership programme.
He told conference delegates: "Some research suggests moving along these lines hasn't been as effective as we thought."
He warned teachers to check initiatives for real evidence. Building a learning process around the results of a questionnaire that took five minutes to fill in would not stand up to scrutiny. But using different teaching methods was still valid and worthwhile "because anything that encourages variety must be a good thing".
Such methods also put learners at the centre, he added.