What not to do with pupil performance data

Ofsted has warned against excessive use of data to assess pupil performance because it is not always reliable. Here are 5 common data practices that you could stop doing, and strategies to use instead, to identify what your pupils know and drive down teacher workload.

Last reviewed on 5 November 2018
School types: All · School phases: All
Ref: 35406
Contents
  1. Stop trying to make precise predictions of pupil outcomes
  2. Stop using 'flight paths' to track pupil performance
  3. Avoid analysing very small groups at school level
  4. Stop trying to predict progress 8 scores
  5. Don't expect too much of assessment data
  6. Get stakeholders on board

Stop trying to make precise predictions of pupil outcomes

Teachers are often required to predict grades that the pupil 'should' achieve years in the future - either manually or using a software system - but these predictions are rarely accurate enough to make them valid.

Yet, pupils' progress is still measured against their grade predictions in order to:

  • Hold teachers to account
  • Predict school-level performance
  • Inform interventions needed by pupils with the lowest predicted grades, even though there isn't necessarily a common factor as to why these pupils are predicted the lowest grades 

Pupils in this lower group will need different forms of support, and your teachers will be able to tailor interventions to pupils' circumstances using their professional judgement rather than relying on a grade prediction. Take a look at our range of articles on using interventions to raise achievement for help. 

Ask your teachers to predict a two

Read next

Also in "Using data for school improvement"