1. Keep things organised
Ensure there’s a clear sense of structure to your data, so that you can access any information relating to the whole school, followed by each key stage, year group, class, group and pupil.
2. Remember the vital statistics
Whatever system you use, you must be able at any point in the year to show the percentage of pupils below, at, or above the expected standard – as well as those pupils making below expected, expected, or accelerated progress. You should be able to then break this down into groups, such as boys, girls, more able pupils, children with SEND and so on.
3. Use colours
Different shades can indicate degrees of significance, but don’t overdo it. Set viable and commonly accepted thresholds, such as, for example, the point at which a pupil starts making accelerated progress.
4. Include pupil numbers and percentages
Remember that percentages with small cohorts can be misleading, and that some pupils may be counted in data more than once. A disadvantaged boy with SEND who is also a persistent absentee could potentially be affecting your data multiple times. By the same token, that ‘100% of disadvantaged pupils in 6T making more than expected progress’ may represent just one pupil.
5. Tailor your data to different audiences
The same data set is unlikely to serve pupils, parents, teachers, senior leaders, governors, Ofsted and LAs equally well. Everybody likes a summative overview of distilled ‘big message’ data presented over a page or two – the finer detail can sit beneath this.
6. Analyse the data and extract key messages
There’s no point in just presenting a load of numbers – formulate a narrative.
7. Formulate a response
What patterns or performances warrant further action or a strategy? Which are things that you should just monitor and keep an eye on?
8. Ensure your teacher assessments are accurate
Fail to do this, and it’ll be ‘rubbish in, rubbish out’. Get involved in in-school and school-to-school moderation, and perhaps attend some LA training, to ensure accuracy in your teacher assessments.
9. Don’t expect the data to explain everything
Data is useful for quantifying performance, but it can’t tell us everything about the human beings behind the numbers. There may well be one or more ‘back stories’. If you opt to highlight any outliers, make sure you can refer to case studies of actions taken with them.
10. The same advice applies elsewhere
The principles above are equally relevant to other forms of data collection, such as attendance, persistent absence, exclusions, parent surveys and so forth.
Grahame Smith is school improvement manager at Havering School Improvement Services
