Assessment can seem a bit scary. Teachers often worry about getting it right, seeing it as adding to workload and causing stress for kids. But in reality, teaching is all about making assessments.
Every day, every lesson, teachers check the children understand what they’ve just been taught, and make inferences about what has and hasn’t been learned. The only trouble is, we do all this so informally, the quality of our inferences isn’t always as good as it needs to be.
I once heard a well-known educationalist say that if teachers don’t understand the concepts of reliability and validity then they have no business being teachers. I knew what the words meant but had very little idea of how they related to assessment.
Terrified, I went away and started reading up on the subject. The literature is vast and often impenetrable; you could study it for years and still only scratch the surface.
Broadly though, validity is commonly defined as the extent to which a measurement corresponds to the real-world thing you’re trying to measure. In essence, are you measuring the things you claim to be measuring?
Reliability represents the extent to which a measure is consistent; would our assessments produce similar results to someone else’s? Would a student respond similarly on two different days?
A fine performance
We know these things matter when it comes to high-stakes assessments, but what about the kind of assessment we do day in day out? One of the biggest problems is caused by the belief that we can see learning.
In actual fact, all we ever get to see is students’ performances. We see what they do: writing essays, answering questions, solving equations, taking part in sports, etc.
Learning takes places inside students’ brains and we never get to observe it directly. A helpful definition of learning is that it is the retention and transfer of knowledge.
If this is true then it follows that we can’t see whether something has been retained until we wait, and can’t see whether something can be transferred to a new context until we go somewhere else. The bottom line is that we cannot see learning in the here and now. All we can see is performance.
If we ask students whether they can still do what we showed them how to do earlier in the lesson, what do we actually find out? If they can’t answer successfully then we might get some useful feedback about how we need to adjust our teaching, but if they can, this tells us nothing about what they’ll be able to do next lesson.
We draw erroneous conclusions about what our students know and can do all the time, but worse, so do they. Children are excellent at working out what answer the teachers want them to give.
Often, these responses are evidence only of mimicry. If they guess what we want and answer correctly, they are likely to remember that they knew the answer.
But what they won’t keep in their memory is what the answer actually is. This is the illusion of knowledge.
If you ask students, do you remember last Friday’s lesson, they’ll say yes. Of course they remember it. But if you ask them what they remember, they may well not recall anything useful.
The solution is to make our assessments less instant. If we wait until next lesson to ask students what they remember about the last one, we’ll generate more reliable, more valid inferences about what they’ve learned.
A great strategy is to harness the power of retrieval practice by beginning lessons with questions not just about the previous lesson, but last week’s, last month’s and last year’s.
If these questions are multiple choice, responses can be generated very rapidly. Not only do teachers get more reliable and valid information about their students’ progress, so do the young people themselves.
We frequently think we know things we don’t actually know; it’s not until someone asks us that we discover the true extent of our ignorance.
David Didau is an independent education consultant and writer. He blogs at learningspy.co.uk.
Sign up here for your free Brilliant Teacher Box Set