Progress scores for individual classes can’t be used to form accurate judgements about teacher effectiveness.
This is because the composition of classes is never completely random: students are moved in and out of groups depending on a range of factors, including behaviour and recent levels of attainment. This happens in traditional sets, mixed-ability classes and streaming models alike. Over time, students making similar rates of progress tend to get clustered together – it’s this variable that matters most.
Unsurprisingly, classes predominantly formed of students who are broadly compliant and reasonably well-motivated are far more likely to achieve higher progress scores. Assuming basic levels of competence, this has little to do with the quality of teaching. Think of it as a manifestation of the Matthew Effect instead.
Remember that Progress 8 is a measure designed to make judgements about the performance of whole year groups and wider school effectiveness. What it can’t do is measure individual teacher effectiveness, irrespective of whether a set of class results are conveniently good or rather disappointing.
Further Reading 📑
Doug