Principles of Great Assessment: Increasing the Signal and Reducing the Noise

Screenshot 2017-03-09 17.03.29.png

After the government abolished National Curriculum levels, there was a great deal of initial rejoicing from both primary and secondary teachers about the death a flawed system of assessment. Many, including myself, delighted in the freedom afforded to schools to design their own assessment systems anew. At the time I had already been working on a model of assessment for KS3 English – the Elements of Assessment – and believed that the new freedoms were a positive step in improving the use of assessment in schools.

Whilst I still think that the decision to abolish levels was correct, I am no longer quite so sure about the manner and timing in which they were removed. Since picking up responsibility for assessment across the school, I have come to realise just how damaging it was for schools to have to invent their own alternatives to levels without anywhere near enough assessment expertise to do so well. Inevitably, many schools simply recreated levels under a different name, or retreated into the misguided safety of the flight path approach.

I would like to think that our current KS3 assessment model, the Elements of Expectation, has the potential to be a genuine improvement on National Curriculum levels, supporting learning and providing reliable summative feedback on student progress at sensible points in the calender. Even though it is in its third year, however, it is still not quite right. One of the things that I think is holding us back is our lack of assessment literacy. I am probably one of the more informed staff members on assessment, but most of what I know has been self-taught from reading some books and hearing a few people talk.

This year, in an effort to do something about this situation and to finally get our KS3 model closer to what we want, we have run some extensive professional development on assessment. Originally, I had intended to send some colleagues to Evidence Based Education’s inaugural Assessment Academy. It looks superb and represents an excellent opportunity to learn much more about assessment. But when it became clear budget constraints would make this difficult, we decided to set up and run our own in-house version: not as good (obviously) and inevitably rough around the edges, but good enough, I think, for our KS3 Co-ordinators and heads of subjects to develop the expertise they need to improve their use of assessment with our students.

The CPD is iterative and runs throughout the course of the year. So far, we have established a set of assessment principles that we will use to guide the way we design, administer and interpret assessments in the future. In the main, these principles apply to the use of medium to large-scale assessments, where the inferences drawn will be used to inform relatively big decisions, such as proposed intervention, student groupings, predictions, reporting progress, etc. Assessment as a learning event is pretty well understood by most of our teachers and is already a feature of many of our classrooms, so our focus is more on improving the validity and reliability of our summative inferences.

I thought it might be useful and timely to share these principles over a series of posts, especially as a lot of people still seem to be struggling, like us, to create something better and more sustainable than levels. The release of Daisy Christodolou’s book Making Good Progress has undoubtedly been a great and timely help, and I intend it to provide some impetus to our sessions going forward, as we look to implement some of the theory we covered before Christmas into something practical and useful. This excellent little resource from Evidence based Education is an indication of some of the fantastic work out there on improving assessment literacy. I hope I can add a little more in my next few posts.

If we are going to take the time and the trouble to get our students to sit assessments, then we want to make sure that the information is as reliable and valid as possible, and that we don’t try and ask our assessments to do too much. The first in my series of blogs will be on our principles of assessment design, with the other two on ethics and fairness and then, finally, reliability and validity.

All constructive feedback welcome!

Advertisements