Clarifying contradictions in CVA

Since the publication of ASCL guidance paper 26 Understanding the CVA Model, several members have asked for clarification of the adjustments used by the DfES.

In particular, they have asked if it can really be the case that students who are the oldest in their year group have their predicted GCSE grade lowered by 14.2. Members had assumed that the reverse would apply.

The table shown in the guide is accurate. The key point to remember is that CVA is designed to measure typical student progress over the five years from key stage 2 to 4.

When the DfES statistical unit carried out the research, it found that "the differences in outcomes accounted for by age at each key stage diminish as pupils get older". In other words, younger students normally catch up by making faster progress over the five years.

Hence, the DfES says that by the end of KS4 they would expect older students to be progressing more slowly than younger students. If your school deviates from that national trend, with older students making more rapid progress, then you are adding value.

A similar question arose from the weighting for first language other than English (EAL), where EAL students have 23.83 added to their predicted GCSE score.

Once again, the DfES statisticians found that from KS2 to 4 EAL students 'make more progress', ie they catch up on some of their peers as the effects of language/cultural barriers diminish. By adding 23.83 to the predicted score, the CVA model takes account of what typically happens across all schools in the country. If EAL students in your school achieve even higher results than the DfES predict, then you are adding value.

CVA attempts to offer a prediction of pupil progress that takes account of all the factors that might influence a student's academic performance. If, once all of these factors have been identified, your school is still above the CVA prediction, DfES argues that your school can then claim to be having a positive impact upon students.

I hope that the guidance paper and this further commentary will help school leaders to see inside the workings of CVA. By understanding how it operates, we can also make a judgement as to whether it presents an accurate picture. We can then decide if it is necessary to present additional evidence that offers a more balanced view.

By Peter Kent, author of Understanding the CVA Model