In preparation for the recording of my recent podcast with Dylan Wiliam, I took some time to reflect on formative assessment, the issue that Wiliam is perhaps most closely associated with. I realised that there have been two big step-changes in my teaching practice. This first was when I started to understand formative assessment, and the second was when I discovered cognitive load theory and found a better way to structure my teaching. The former occurred largely before I started blogging and so I haven’t written about it much.
Formative assessment is a wide and abstract concept and that’s why I think we all struggle to get a handle on it. Often, it becomes understood through certain practices, such as the use of mini-whiteboards or exit passes. One common misconception is that a formative assessment is a quiz that you give to students. You can certainly use such a quiz formatively but, as Wiliam explains in the podcast, formative assessment is a property of the inferences you draw from the quiz rather than the quiz itself. So that’s the problem right there, we are discussing inferences rather than something more concrete and easy to grasp.
Wiliam’s settled definition of formative assessment is:
“An assessment functions formatively to the extent that evidence about student achievement elicited by the assessment is interpreted and used to make decisions about the next steps in instruction that are likely to be better, or better founded, than the decisions that would have been taken in the absence of that evidence.”
This does not mean that we plan for formative assessment and leave a blank space afterwards to be filled only once we have the evidence from the assessment. I have heard people argue that it is impossible to plan three lessons in advance because the third lesson would need to take into account formative assessment evidence from the first one. That may be a way of enacting formative assessment, but it is definitely not a requirement and would strike me as a particularly stressful way to organise your teaching. This view is perhaps based in thinking far too big when the stuff that makes the difference is those little details.
So I thought I would outline how I have enacted formative assessment in my own maths and physics teaching.
1. In the moment
Once I have demonstrated a worked example, I ask students to complete a virtually identical one on their mini-whiteboards and hold this up. Every student is expected to do this with no opt outs. If there are multiple stages to the worked example then they will complete each stage and hold this up for checking before we move to the next stage. This will involve the teacher in correcting mistakes and misconceptions and sometimes re-explaining. We call this doing ‘slow motion problems‘. It is far better to haul someone back on the bus than to reach the destination and find some passengers never made it on to the bus at all.
The lessons are structured around a PowerPoint with explanations, multiple worked examples and questions for students to complete. There are usually too many to complete within the lesson time and part of the teacher’s role is to modulate progress through the materials according to the evidence they see on the mini-whiteboards.
2. Lesson by lesson
Each lesson begins with a starter activity that students are expected to work on unprompted. Usually, these tie in with the previous lesson. However, some are strategically placed to practice something addressed a few weeks prior and some are there to remind students of a concept that will be useful in the upcoming lesson. Students complete this starter, often on mini-whiteboards and this prompts some review and reteaching.
Each week, one of these starters is a ‘question-of-the-week’, an example of spaced practice that is intended to be displaced by about three weeks from the initial teaching. These are self-assessed by the students and collected by the teacher. The results of these questions are logged and compared between classes – I always work in teams of parallel classes – to see if one class has been more successful than the others.
I also draw a box on the board at the start of every lesson where students can write the numbers of any questions from the homework that they struggled with. I usually try to address these in some way before the end of the lesson, either with the whole class or a subgroup of the class while other students are working independently. This was one of the formative assessment strategies I adopted first in my teaching and I have found it to be extremely powerful in exposing things that I haven’t taught very well but would otherwise assume I have adequately covered.
3. Formal assessments
Every three weeks or so, students will have a formal assessment. Although we compare class data for question-of-the-week, it is these formal assessments that provide the richest data and that sit at the sweetest spot for using that data. All teachers of the parallel classes use the same set of teaching materials, starters, homeworks etc. You can move from one class to the other and see closely equivalent lessons. Paradoxically, this means any variation we do find in performance is more easy to attribute to specific things.
We may see, for example, that one class has performed better on question 1c that the other classes. Once we talk to the teacher and ask them to demonstrate how they would teach that question if it were a worked example, we usually find that they have added an additional step or bit of explanation. For instance, this process has led me to recently adopt a strategy for chain-rule differentiation questions where I draw a table at the side of the question. I would not have thought of this myself. The table strategy has now been written into the teaching materials for future cohorts and into the revision materials for the current cohort prior to the end-of-semester exam.
The reason end-of-topic assessments work so well for this process is that everyone can still clearly remember teaching the content. This is less effective for spaced practice, although spaced practice does highlight gaps to be addressed.
As mentioned, we also have end-of-semester exams. However, these cover such a large body of work that it is harder to isolate those specific strategies. Instead, they mainly inform us on areas of strength or weakness. They have prompted us, for instance, to completely review some units of work.
Formative assessment in practice
You may or may not care for the strategies I have outlined. You may have better ones or approaches better suited to your subject. That is fine. Formative assessment is not really about the strategies but it cannot be adequately explained without reference to those strategies. It is worth considering three tests when planning for formative assessment: Will a specific strategy elicit evidence from all of the students? Do I have strategies in place for each of the short, middle and long terms? Will a given strategy give me evidence of a specific issue that I can do something about?
An example of something that can fail the final test is asking students to produce a complex product such as an essay without having a series of smaller checks in place beforehand. If a student writes a poor essay, it can be hard to tell whether this is about particular writing skills or subject knowledge, because they are all conflated. However, if you have already assessed subject knowledge, sentence construction etc., you can start to narrow this down.
Formative assessment is perhaps not as exciting a subject for educationalists as jobs of the future that don’t exist yet or grand political narratives, but it is at the level of prosaic teacher craft knowledge that we will, collectively, improve our profession and become more effective. Formative assessment sits right at the heart of that craft knowledge.