The streaming of children at primary school by their ability is actually widening the achievement gap between low and high-attaining pupils. Our research has shown that streaming in primary school leads to higher marks for children in the top class, but lower marks for those in the middle and lower classes compared to those in mixed ability classes.
Streaming – when children are placed in a class based on general ability and taught in that class for the whole time – is on the increase among five to seven-year-olds in UK primary schools. While around 2-3% of year 2 pupils were streamed in the 1990s, by 2007 this had increased to 17% in England, 16% in Scotland, 20% in Wales and 11% in Northern Ireland.
Streaming for six year olds
Our research, presented at the British Educational Research Association conference, used data from the Millennium Cohort Study (MCS) to investigate how grouping children by their ability into upper, middle and lower streams in year 2 (six to seven-years-old) relates to their academic progress. We also took into account other key child, family and school characteristics.
The MCS follows the lives of around 19,000 children born in the UK in 2000 to 2001. Five surveys of cohort members have been carried out so far – at nine months, three, five, seven and 11 years.
Our research focused on children in families in England where information on streaming had been provided by their primary school teacher – and we had data from the National Pupil Database on their test results at Key Stage 1 (five to seven-years-old). The analysis was based on 2,544 children in 307 primary schools, of whom 83% (2,098) were not streamed, 8% (222) were in the “top” stream, 5% (130) were in the “middle” stream and 4% (94) were in the “bottom” stream.
Top streams do better
The graph below shows the average point scores in reading, writing, maths and science for children by stream and for the majority of children who were not streamed. The findings demonstrate that seven-year-olds placed in the top stream have the highest scores. Those in the middle or bottom streams are doing less well overall, and in reading in particular.
From here we took account of a wide range of child, family and school characteristics. We assessed whether this association with scores in Key Stage 1 was largely due to, for example, the child’s earlier academic ability, their gender, how educated their parents were or the school intake.
Children’s characteristics, such as their previous school performance, gender, age and health, had the greatest impact on mitigating the association between streaming and academic performance.
We found that regardless of the child’s characteristics, in comparison to the majority of children taught in mixed ability classes, those in the top stream were likely to have better overall marks, while those in the middle or bottom streams were likely to have lower average marks.
Neither socio-economic factors such as parental education, income and health, nor other factors such as a child’s enjoyment of school and friendships had little impact on the relationship between their stream placement and Key Stage 1 results.
This meant that when all child, family and school characteristics were included in the analysis, the relationship between a child’s stream and his or her scores remained very significant. Compared to the majority of children in mixed ability classes, those in the top stream had higher average scores and children in the middle or bottom stream had lower average scores. The relationship only lost its statistical significance for maths scores for children in the middle stream.
Family time helps
A child’s earlier academic performance was the most significant positive predictor of higher scores at age seven, whereas behaviour difficulties had the opposite effect. Being born in the autumn (meaning the child was older for their year), was also associated with greater academic progress in maths and overall performance at Key Stage 1.
Joint family activities were associated with higher scores across the subjects, as was how much interest parents showed in their child’s education.
In many ways the findings are counter-intuitive, given that matching work to pupils’ current level of attainment should enhance learning. The most likely explanation as to why this is not the case is that grouping children by ability changes teachers’ expectations.
This impacts on what is taught to different groups, how it is taught and the unspoken messages given to pupils. Schools need to be aware of these issues when making decisions about structured ability grouping.