Children as young as three are being grouped by "ability" to prepare for Sats, research has found.
Teachers say they are expected to group children in order to prepare them for the phonics check taken at the end of Year 1 and the KS1 Sats taken at the end of Year 2, a study by the UCL Institute of Education has found.
The study, commissioned by the National Education Union (NEU), discovered that, while 45 per cent of teachers think grouping damages some children’s self-esteem, it is seen as a “necessary evil” due to the pressure that schools are under from the statutory tests.
“The findings make for challenging reading,” Mary Bousted, joint general secretary of the NEU, said.
“It’s an absolute disgrace that the pressure on schools to ensure pupils pass tests means children as young as three consider themselves ‘low ability’ right at the start of their academic life, a belief that could impact on their self-esteem, carry on throughout their schooling and determine the direction of their adult lives.
“The National Education Union will work hard to lobby the government to address accountability and curriculum pressures that lead to the labelling of children. As a first step to improving the situation, the government should commit to make the phonics screening check non-statutory.”
The research has been commissioned as the government prepares to reintroduce a Reception baseline assessment, which could mean teachers having to test children at the beginning of Reception in literacy, maths, communication and possibly "self-regulation".
“There is a real risk that if bring in an assessment at the beginning of Reception that encourages further labelling of children,” said the UCL Institute of Education's Alice Bradbury, one of the co-authors of the report.
More than 1,400 teachers were involved in the study, including 1,373 who responded to a survey and the others in four focus groups and four case study schools.
The survey found that across nursery, Reception and key stage 1 teachers most commonly grouped children in phonics classes (76 per cent), then maths (62 per cent), reading (57 per cent) and literacy (54 per cent).
Even in nursery classes, more than half (58 per cent) of children were grouped for phonics sessions.
The survey found that 52 per cent of teacher felt grouping did improve results. But there were also fears that it could reduce children’s confidence and limit their expectations.
"It was deemed an efficient way to plan differentiated lessons but it made children very competitive and destroyed self-esteem in some pupils," one teacher told the researchers.
Beatrice Merrick, chief executive of early years association Early Education, said: "Many, including parents, will be shocked by the pervasiveness of setting, streaming and other 'ability' based interventions with children aged 3 and 4, and even more so given the evidence that shows this harms far more children than it helps.
“It is worrying that so many teachers feel pressured into such steps against their better judgement; we hope this will lead to dialogue in all schools about how best to ensure all children achieve their potential without fear of early labelling or artificial constraints on their learning."
Minister for Children and Families, Robert Goodwill, said: “Teachers and early years staff are best placed to make decisions about the teaching methods they use. There is no statutory requirement that suggests children should be grouped by ability.
“We are clear that, while assessment is a fundamental part of children’s education to measure progress, it should not cause significant stress or anxiety.”
Want to keep up with the latest education news and opinion? Follow Tes on Twitter and Instagram, and like Tes on Facebook.