Choosing which of your colleagues will carry out judging, how many judgements to make, and signing the judges up

Choosing your judges

One of the benefits of Comparative Judgement is that it minimises individual judge bias if the judging is shared. With a large pool of judges, each script is likely to be seen by many judges, so the bias of any one judge is balanced by the views of other judges. In other words, the more judges you have, the less liable your assessment is to bias.

So, although you can judge alone, if you can share your judging you will minimise bias.

We think it is ideal if all staff, including Teaching Assistants if appropriate, are involved in the judging, particularly when the school takes up the system for the first time. This way they will all get to understand how the system works, how simple and efficient the process is, and appreciate where the children's scores come from.

You could organise for all staff to attend a judging session, all bringing along their own laptops to do their own judgements, and then perhaps have a discussion about how it all went afterwards. This way you should get maximum buy-in from your colleagues, and they ought to be keen to take part in future judging sessions.

In reality, we find that although many schools do this, there are also many who use just a core set of judges each time, or choose to select those closest to those year groups for teaching purposes, or just see whoever is available and willing. You can involve different sets of judges for different year groups' tasks. It's all down to how it best suits your colleagues, the logistics, your timetabling, etc.

Choosing the number of judgements for each judge

The more judgements that are made by judges, the more reliable (precise) your results will be.

As a general rule, we would suggest carrying out 10 judgements per candidate for our national judging sessions, spread across all of your judges.

For national tasks (as opposed to in-house tasks), you can not change the Candidate Judgements value from 10, and you can not change the Expected Candidates value.

You can only change the Expected Judges value.

For example, if you have 60 candidates to judge, you would need around 600 judgements.

You would divide this number by the number of judges you have, to find the required number judgements per judge. So if you have 12 judges, this would mean 600/12 = 50 judgements per judge. Or if you have 10 judges, this would mean 600/10 = 60 judgements per judge, and so on.