One of the key principles of a good Ofsted inspection is that when the final meeting occurs with a Head of Department, subject leader, or Principal, there should be no surprises. The last thing anybody wants is for the inspector to say "Everything is good except that you haven't addressed X", only for the interviewee to say "Yes we do, but you didn't ask for that information."
Had the inspector said the day before, "I'm starting to get the impression that...", the teacher or Principal could have responded by saying "You're wrong", or even "You're right" – and then start the process of addressing the matter before that final meeting.
What we're talking about here is not simply data or information, but timely data. The same applies to student assessment. Telling a student that they're not doing so well is rather pointless if you wait until the day before the exam – what are they expected to be able to do about it?
And the same thing applies to whole school data or, in the case of Local Authorities and Multi-Academy Trusts, the data from several schools. In order for the data to be of any practical use, it has to be timely.
Unfortunately, although many articles on leadership declare that data collection should be timely, very few of them define what the word 'timely' means. The closest any of them come to it is an article entitled Six Steps for School Leaders to Use Data Effectively. In the article, the author Chris Balow states that:
"The data system must make the data timely and accessible to ensure that the data is gathered and put to use."
That phrase 'put to use' is the nub of it. You need to collect the data in early enough to be able to use it for decision-making. So this begs the question: what counts as 'early', and what counts as 'too late'?
The answer will vary between institutions of course, but a useful approach is to work backwards: when will you want to start making decisions based on the data you collect?
Using that as a starting point is quite helpful. The summer term, you will realise, is too late if staffing and curriculum changes are justified. For example, if you decide to put a teacher from school A into less effective school B for the start of the next academic year, they will need to know before the summer term so that timetables can be amended accordingly and lesson preparation and school visits and meetings to be planned for.
If the decision is likely to involve shifting financial resources around, probably the start of the spring term will be too late, because that's when new budgets tend to be set.
Because of these conflicting priorities, ie conflicting in terms of their timings, collecting in data 'early' really comes to down to collecting data as frequently as possible without the process becoming so burdensome that it has a negative impact on school efficiency and outcomes.
An iterative process is quite useful in this context. Collecting the data once a year is never likely to be timely. On the other hand, collecting it every day will be a prime example of the data tail wagging the school dog.
In order to decide when to collect the data, consider the nature of the data and what you will be able to do with it. For example, there is no reason that attendance and lateness data cannot be collected on a weekly basis. Doing that would, over a period of half a term, give you enough data to see that school A is doing very well but school B is not – and then to do something about it in the following half-term.
On the other hand, summative assessment data could be collected at the end of each half-term. A weekly assessment of pupils might be considered overkill, but a half-termly one is both manageable and useful – because by the second (ie end of term) assessments any underlying differences between schools should start to become clear.
In summary, the following principles apply:
Decide what kind of decisions you want to be able to take...
... And when it will be necessary to take them.
Data should be collected early enough for it to be acted upon.
What counts as 'early enough' will vary according to the nature of the data.
And don't forget that the process of actually collecting the data can be taken care of easily using Groupcall's Xporter.