There is a "weak relationship" between Year 2 teacher assessments and Year 3 comparative judgement scaled scores, a study has shown.
The research shows that using comparative judgement for assessment produces more reliable results, according to the group behind the research.
More than 1,300 teachers from 120 schools took part in the study. They judged the written work of Year 3 pupils by comparing two pieces of work and deciding which one was better.
Researchers compared the Year 3 comparative judgement average scaled scores with the end of Year 2 statutory teacher assessment results in 21 schools. They discovered a “weak relationship" between the two sets of results – despite the fact they were carried out only a few months apart.
The study, which looked at more than 6,200 pupils’ work, also finds that teachers who abandon marking criterion to use comparative judgement show a high degree of consistency.
The overall reliability of the judgements was 0.91 out of 1.0 – which roughly equates to a margin of error of +/-1 mark on a 20-mark question.
The research was conducted by No More Marking – an online engine to help teachers with comparative judgement assessments. It finds the results are more “consistent” than traditional teacher assessment because teachers judge the work of pupils from other schools anonymously.
Daisy Christodoulou, No More Marking's director of education, told Tes: "You would expect quite a tight fit between those two, because ultimately they are measuring writing. But what we are finding is that there is quite a weak relationship."
She argues the "mismatch" occurs because comparative judgement is “more reliable” than teacher assessment.
Ms Christodoulou said: "The issue is that every school is placing these national boundaries in a different place. It is internally consistent within their school, but it is not consistent outside it."