Key points
- It’s critically important to show students what success looks like
- In most cases, success criteria fail to do this
- While success criteria might be relatively simple to implement, it is not a high yield strategy when compared to other practices
- Time spent writing and presenting success criteria is better spent planning examples, models and checks for understanding
Why success criteria fail
In classrooms across Australia, teachers are working hard to show students what success in learning looks like. For all kinds of reasons, this is good. What is less good is the pervasive idea that presenting success criteria at the start of a lesson is an effective way to do this.
I am unconvinced.
What are success criteria?
Success criteria are descriptors of successful performance, linked to a learning goal. In almost all cases, they are presented (and discussed) at the beginning of a lesson. In relation to a precise learning intention, success criteria aim to help students answer the question: “How will I know when I’m being successful?”
Let’s look at an example.
Learning intention
We are learning to construct sentences with the subordinating conjunctions before and after
Success criteria
- I can use before and after to show the time relationship between two actions or events
- I can write accurate sentences that begin with before and after
While success criteria seem useful on the surface, in many cases they create more problems than they seek to solve. In this post, I explore three problems with success criteria, outlining practices that may represent a better use of teachers’ time.
Success criteria don’t show students what success looks like
The most obvious problem with success criteria is that they don’t show students what success looks like! It’s important to be clear on this. Performance descriptors, no matter how well worded, can never adequately convey quality because they are abstractions.
Success criteria make reasonable sense to the expert (who already has the schema needed to conceptualise quality), but to the novice, they remain impenetrable. (If you want to read more about the limitations of performance descriptors, read Making Good Progress by Daisy Christodoulou.)
Many success criteria are redundant
Success criteria are often a re-hashing of a perfectly good learning intention. When learning intentions and success criteria are almost the same (e.g. a learning intention is rephrased in “I can” statements), we introduce what cognitive load theory describes as the redundancy effect (Sweller, 1988). This is the result of asking students to process multiple versions of similar information for no conceptual gain.
Time is desperately needed to work on higher-impact things
Time spent writing and presenting success criteria is time not spent developing shared models, examples and checks for understanding; these are the kinds of practices that are much more likely to convey the criteria for quality in learning (e.g. see Kirschner, Sweller & Clark, 2006; Rosenshine, 2012). Somehow, we find ourselves in the curious position of having schools that are drowning in success criteria, and yet have not planned checks for understanding.
We have muddled priorities.
So why are sucess criteria so popular?
At the risk of sounding cynical, I suspect the practice of sharing success criteria is popular because it’s visible, quick to implement, and provides a veneer of “consistency” across classrooms. If we walk the halls to see every teacher starting the lesson with success criteria, then all students will know what success looks like, right? No. If communicating quality across the curriculum were that simple, teaching would be easy.
So what practices should we prioritise instead?
Before we spend any time on success criteria, we should ensure teachers are given time to:
- Collaboratively plan the representations, models and examples that will make clear what “great” looks like in any given sequence of learning
- Create plentiful opportunities for live, dynamic modelling (paired with precise think-alouds)
- Prepare varied examples, including worked examples, non-examples and boundary cases (what Dylan Wiliam and Siobhan Leahy might call “near miss examples”, in which students are shown explicitly examples of what “not quite” looks or sounds like)
- Engage students in simple comparative judgment to explore different aspects of quality (“Let’s compare topic sentence A and B, this time focusing on the choice of language.” Which do you think is better, and why?)
- Check all students’ understanding frequently, and flood the environment with direct feedback (leaving students in no doubt about what constitutes quality)
Wrapping up
I think it’s time for a more critical conversation about the value of success criteria in teaching. I hope the reflective prompts support a continuing critical conversation in your school and as ever, I’d love to hear from you if you have a perspective to share.
Staffroom reflections
- Why are we currently implementing success criteria? What do they achieve?
- What are the dangers in presenting success criteria at the start of a lesson, before we have shown students models or examples?
- What are some of the limitations of success criteria?
- Are there any sequences of learning in which we need to focus on building more (or better) models, examples and non-examples?
