I’m working on a guide to formative assessment for classroom teachers. I want to do two things: first, set out clear principles and introduce the research underpinning them; second, show how these principles work in practice through discussing an extensive range of examples.
Why?
Two millstones burden formative assessment; two current trends demand we look at it afresh.
Millstone #1
Effective formative assessment remains rare. The Assessment Commission Report found that:
Formative classroom assessment was not always being used as an integral part of effective teaching”;
and a focus on National Curriculum Levels in schools meant that:
Instead of using classroom assessments to identify strengths and gaps in pupils’ knowledge and understanding of the programmes of study, some teachers were simply tracking pupils’ progress towards target levels.”
The Carter Review found that:
The most significant improvements [to Initial Teacher Training] are needed for training in assessment… there are significant gaps in both the capacity of schools and ITT providers in the theoretical and technical aspects of assessment.”
Millstone #2
Meanwhile, some of our most thoughtful teachers have turned their backs on formative assessment. David Didau has posed valid concerns about the the conflation of learning and performance, and some formative assessment techniques; rather than refining formative assessment however, he uses these concerns to:
Conclude that the ‘big idea’ of AfL is wrong.”*
Meanwhile, Joe Kirby’s post on workload ‘hornets’ rejects key tenets of formative assessment out of hand:
No writing, sharing or copying learning objectives or outcomes…. No mini-plenaries or checks on progress within a lesson.”
Fresh attention
Conversely, two current concerns in English schools demand that we reexamine and refresh formative assessment:
- Formative assessment needs to be adapted to suit the demands of a knowledge-rich curriculum. The underpinning research and most of the principles rest unchanged; the details and the techniques may differ in nature or emphasis from those which dominated my early experiences of formative assessment.
- Cognitive science offers us a much deeper understanding of why formative assessment works; this understanding can guide our refinement of existing techniques.
Seven principles of formative assessment
I believe formative assessment can be cast in seven principles:
- Determine exactly what students need to know and be able to do.
- Align every aspect of lessons to their purpose.
- Show students what’s expected.
- Respond to students’ understanding between lessons.
- Respond to students’ understanding within lessons.
-
Provide feedback which causes improvement.
-
Guide refinement.
The principles are hierarchical: successfully showing students what’s expected relies on having determined exactly what they need to know and do; providing feedback which causes improvement depends on a good understanding of what students have understood. (There’s a clear debt here to the careful sequencing of teacher improvement in leverage coaching). I believe these principles apply across subjects and ages (if you disagree, now’s your chance).
This looks very different from the canonical formulation of formative assessment as five strategies:
My view is that a hierarchical formulation sets out priorities more helpfully, and where the bulk of responsibility lies – with the teacher – more clearly. Abundant scope remains for peer and self assessment once students know what they’re expected to do and have sufficient guidance to succeed.
A direct way to apply each principle
In the guide, I intend to exemplify each principle through detailed discussion of one technique. I plan to use the following techniques (links lead to previous, somewhat dated, posts about them):
- Determine exactly what students need to know and be able to do: setting clear learning objectives
- Align every aspect of lessons: Ruthless application of a checklist (do you have a better suggestion?)
- Show students what’s expected: sharing lesson objectives and model work
- Respond to students’ understanding between lessons: exit tickets
- Respond to students’ understanding within lessons: hinge questions; questioning and classroom discussion
-
Provide feedback which causes improvement: RAG/dot marking
-
Guide refinement: redrafting
(There’s more to each principle than one technique; this is primarily a way to convey the principle through a detailed example. I will discuss a range of other approaches briefly).
Now, your turn:
- Are these principles well-chosen and well-formulated? Why (or why not)?
- Are these techniques good examples of the principles?
- What examples of these principles in practice can you share?
- Have you tried to adopt any of these principles or techniques? Would you be interested in describing how it went?
If you have answers to any of these questions, or are interested in commenting on a draft chapter of the guide, let me know.
* I’ve since critiqued David’s arguments in more depth here.
Well written and makes the point of AfL clearer.
Still, the first (and most important) point: “1.Determine exactly what students need to know and be able to do: setting clear learning objectives” is a place for discussion.
You wrote that cognitive science offers us a much deeper understanding of why formative assessment works. So it is and now there is quite a lot of evidence, that a learner just doesn’t have to know what (s)he needs to know and be able to do, but the learner has to be motivated to know and to do those things (and not just because they are stated in curricula or proposed by the teacher). So the learners have to participate in point 1.
You also cited: “Instead of using classroom assessments to identify strengths and gaps in pupils’ knowledge and understanding of the programmes of study, some teachers were simply tracking pupils’ progress towards target levels.”
While using these steps, it is aslo important to not forget, that each student has different starting point and therefore there cannot be the same objectives to everyone. There could be the same goals (like becoming better in oral presentation or writing in a forign language or stronger or better in analyzing data, etc), but the objectives have to depend on the current level of the learner too.
Thanks for the comment and critique. I agree that sharing the ‘why’ alongside the ‘what’ is important, although I’d add that the evidence strongly suggests that successful learning causes greater motivation more so than motivation causes successful learning.
I also agree that differentiated objectives are important, within the same overall goals: this is what I’ve always used in formulating my own objectives and what I advocate in training teachers.
Joe Kirby’s rejection of mini plenaries or checks within a lesson is not quite the rejection of formative assessment it appears at first glance. If memory serves correctly, he is an extremely strong advocate of using the visualiser and a very deep level of analysis of one or two examples of student work mid lesson with the rest of the class then editing their work accordingly. If you accept for a minute that the whole point of doing the mini plenary or check for understanding is to gather data to inform the direction of future teaching then isn’t Joe effectively doing the exact same thing in his own way?
Fair correction. I’ve always taken Joe’s statements on learning objectives to be a rejection of the underlying principles, and his denial of ‘checks on progress’ look similar to me. It would be interesting to know whether he sees Michaela’s practices as conforming to these principles of formative assessment in more efficient or effective ways. If so, this would be a nice illustration of the weight we need to place on the principles…
Harry: I think your principles are helpful if you want to provide a step-by-step guide for how to implement formative assessment. The danger is that there will be examples of good formative assessment that do not fit your principles. Or, to put it another way, I think what you are proposing is useful as a framework for novices, but it will be constraining for those who are more expert. The five strategy framework that I developed with Marnie Thompson was intended to be inclusive and comprehensive, so that anything that could reasonably be called formative assessment would fit in the model. Our models are different because they serve different purposes. I don’t see any conflict.
Turning to your other points, I don’t think that there is any problem in applying our five-strategy framework to any view of knowledge. I have always maintained that any theory of formative assessment has to be independent of what is to be learned, and independent of any view about what happens when learning takes place. And I do not think that you need to worry about the criticisms posed by David Didau. As I have pointed out in response to his various posts, his criticisms might have some validity but for the substantial and growing evidence that when teachers pay attention to formative assessment, their students learn more.
I see Dylan Wiliam has replied about the five point framework so I won’t restate the point he makes but as it is something that we’ve been working on this year, I’d like to give an example or two of the flexibility of the five point framework and how it works well.
I’m sure you’re aware Harry of the learning scientists great work popularising ideas to best structure practice through things like retrieval practice, elaborative practice, dual encoding etc. The five point framework is easily flexible enough to incorporate them (stage two designing activities/tasks for learning). So without confusing people with another ‘new initiative’ the five point framework is easily flexible enough to incorporate ideas that work.
One other example of this flexibility, and how you can use the framework is some work our science dept have been doing on adapting Mazur’s peer instruction to the secondary classroom. Peer instruction has a wealth of evidence behind it at tertiary level and surprise surprise the in class workflow fits virtually exactly to aspects of the five point structure with virtually no alterations. Eg: the conceptest fits perfectly into stage 2 and 3 where you gather feedback from the students to inform instructional devisions while the turn to your neighbour and discuss aspect is a perfect example of activating students as learning resources for each other.
Or to put it another way, as you are in a position of training teachers I’d stick with the five point framework for now. It really is very flexible while at the same time providing a shared language for design of learning.
Another fair and helpful comment. I’ll have to consider whether I need to restructure the book!
Thanks for the comment and suggestions Dylan. Most of what I’ve learned about conveying formative assessment to others has been through training new or early-career teachers, so it makes sense that this framework is more suitable for novices. My experience suggests that most teachers would benefit from support with at least some of these principles however.
In arguing that knowledge-rich curricula demanded a refreshed approach to formative assessment, I don’t mean to criticise the framework itself, but to echo Daisy’s argument that we need to target our assessments more carefully towards smaller chunks of knowledge and skill. I do think this demands some tweaks to the way we approach formulating lesson objectives (as compared to that advocated in Embedded Formative Assessment).
Your principles for formative assessment are well formulated and display the path to learning that should take place with every lesson. I would add a critically essential principle as #2: Decide what type(s) of THINKING and PROBLEM SOLVING your students must be able to do in order to master the know and do of the learning
objective. Then, as the assessment of the learning objective (exit ticket) is created, the teacher must use this criteria: Will this assessment require students to apply the thinking and problem solving skills needed to master the lesson objective?
At that point, the teacher is THEN ready to plan the meat of the lesson. What modeling needs to occur? What exemplars and rubrics will students need to FEED FORWARD the success criteria for the learning objective. Then as the teacher makes his/her brain transparent by modeling the process of thinking and problem solving, gradual release of thinking and problem solving tasks for students will occur. During this whole process, all participants in the learning cycle have concrete evidence (exemplars, rubrics) to use to FEEDBACK to each other ways to refine their thinking and their problem solving skills to lead to mastery of the know and do for today.
Our children deserve for us to raise the rigor today and every day so they will be ready to tackle and master the thinking and problem solving tasks that are certain to embody every minute of their tomorrow.
Hi Sarah,
Thanks very much for the comment, which I’ve spent a good deal of time thinking over. Your point about ‘deciding what types of thinking and problem solving’ students need to be able to do really underlines the importance of deep curricular thinking before and alongside the planning of assessment. I’m not yet sure how to incorporate this within the guide, but formative assessment will achieve little without it.
Within the planning of the lesson I’d hope that this kind of thinking is implicit in the first and second principles: to establish what students need to know and do, and ensure activities in the lesson are aligned to this, I would say you have to know the kind of thinking and problem-solving it demands. What I’m still thinking through is how best to make this clear!
I’m grateful to you for stretching my thoughts!
Another thing occurs Harry, in terms of how the 5 point formative assessment framework is quite a useful ‘theory of everything’. Assuming you accept the premise that in training novice teachers we want to reduce cognitive load as much as possible when introducing new ideas then the 5 point framework helps to do that. So say for example I was introducing cognitive load theory to trainee teachers (like what I’ve done there!) I could probably reduce cognitive load by tacking it to the 5 point formative assessment framework (stage 2 design of activities/tasks for learning etc). Perhaps, rather than redo the framework it might be more valuable if you made an editable wiki style version and introduced it right at the start of your training. Then as new ideas were introduced or as experiences from teaching practice were reflected on you and your trainees could keep referring back to it as an overarching framework and filling up each stage with examples, related concepts etc etc.
Thanks very much for your post, it’s very thought provoking
I like the idea of an editable wiki and adding things in – but I also feel there’s merit in reducing cognitive load even further by inviting teachers to focus on one point at a time, and in order…
I’ve been working on the guide this week and I’ll confess that, with your points in mind, I have been very tempted to scrap the current structure and go back to the traditional approach. I’ll let you know when I decide. Thanks again for taking the time to argue this out!
Reblogged this on The Echo Chamber.