How often do government departments have to reissue press notices? Following intervention from the Office for Statistics Regulation, the DfE have been placed in that position. The OSR letter can be read at
https://www.statisticsauthority.gov.uk/wp-content/uploads/2017/09/DfE-statistics-Ed-Humpherson-to-Mike-Jones.pdf The revised press notice and the other issue about MAT transfers raised by OSR concern matters dear to the policy objectives of Ministers, so any potentially misleading data are of concern.
However, the DfE statisticians also have to battle with others that don’t always provide data that is of top quality. As reported in an earlier post, about local authority expenditure per pupil, there are a couple of local authorities in one table in that Statistical First Release where the data must be suspect because of the reported level: it both cases, way too low.
Then there is the case of under-reporting by schools in areas such as fixed term exclusions and more specifically for the number of pupils placed on reduced timetables, but not excluded. This is an area where more work is needed to discover what is actually happening, not least in the academy sector. This work is important because of the potential safeguarding aspect.
Local authorities and the local Safeguarding Board may not be in full possession of the facts if academies do not fully report to the DfE. It would be a simple change to add to the funding letter that academies are required to report all statistics via the local authority where they are located unless the Regional School Commissioner has explicitly provided for an alternative system that is as rigorous. At present, this is an issue with one part of the dual system not working as well as the other and creating potential risks for young people.
At least these days, as with the re-issued DfE Press notice mistakes can be rectified when noticed. In the former days before the internet such mistakes could become set in stone. One of my first communications with government was to point out that pupil teacher ratios provided in a written parliamentary answer and reported in Hansard were wrong. I think the first local authority in the list missed the PTR and was allocated that of the next by mistake. It wasn’t picked up before printing and went into the record. An error notice appeared later, but who checks written PQs for later revisions? Nobody, I would hazard a guess. As a result, anyone using that data source would have inaccurate data. It doesn’t matter now, but might have then. One year, the Department had to re-issue a whole Statistical Volume because of the number of printer’s errors.
Today, the record can be set straight quickly and easily, even if the original error is retained as well.
Statistics are important as a source of information under-pinning decision making and debate, hence the need for accuracy. The question of management information that is separate from statistics is one that has always interested me. In some areas, such as the labour market for teachers, I have always believed up to the minute information is important to spot changes in trends as early as possible. However, this data is often in a raw state and not 100% accurate. Where to draw the line between management information and statistics is an interesting and ever changing debate as technology provides ever more exciting tools for data collection and analysis.