As you now calculate prior attainment groups for children who were post levels at KS1, could this feed into VA calculator/ VA progress column on tables so we can get an idea of value added for children based on scaled scores to date. I have a manual version based on my own stab in the dark which highlights children who are massively -ve based on the scaled scores they are achieving within that year group but would be great to do it within insight rather than duplicating data entry.
We only create broad High/Mid/Low KS1 attainment groups. The VA calculation uses much finer-grained prior attainment groups, 24 in total. We’re not going to be able to match that. I feel that we already went out on a limb by calculating the broad groups. I’d worry about pushing it any further with predictions of VA.
If/when the government announce any progress measures from post-levels KS1 data, then we will certainly look to update Insight.
I echo what Andrew says. VA is based on 24 prior attainment groups, and that is likely to change in future with pupils with new KS1 data filtering through to KS2 (current Y5 and below). I always say that in order to predict VA progress measures, we need to be able to do 3 things:
predict pupils’ KS2 scaled scores. We can use practice SATS in Y6 but these will be based on previous year’s test so it’s risky. And we can’t use standardised tests (eg PUMA/PIRA, NFER, GL) because that is a completely different scoring system to KS2 scaled scores.
predict the estimated outcomes for each PAG. We can use previous year’s estimates from the VA calculator as a guide, but only really with any degree of confidence for current Y6 who have levels. We have no published methodology for pupils with new KS1 data and won’t have until autumn 2020.
predict changes to methodology. We don’t even know if the DfE will maintain a combined subject approach for KS1 baseline. They probably will but this will require retrospectively assigning nominal scores to the new KS1 data, and may or may not involve double-weighting maths as now.
As Andrew says, we have developed the broad LMH prior attainment bands using our own method, for reporting purposes, in the absence of any official DfE method. We felt that this was useful for presenting data to governors and external bodies including Ofsted. Calculating VA is a very different thing. In fact Ofsted released a statement last year to that effect: they will not be requesting predictions of progress scores because it’s too difficult to do with any degree of confidence.
Basically, we would not want to provide schools with any data that turned out to be erroneous and may put them in a difficult position. It’s too risky. Using an Insight table to calculate the average change in standardised score between two points, would be a useful, alternative VA measure.
Thanks both. Completely understand and agree.