As the first run of key stage 2 tests approaches, Pete Strauss questions whether we should be thinking in terms of 'levels' at all. The problem with the debate about assessment at key stages 1 and 2 is that it hasn't really happened yet. The most important questions haven't been asked. Consequently, we've ended up with, at best, a sterile and superficial argument about "pencil-and-paper tests" versus teacher assessment, and at worst, an irrelevant row about "workload", external markers and extra supply cover for tests.
Last year I decided that I wanted to find out much more about what the children in my Year 4 class were learning. I spent some time experimenting with different approaches to assessment. These included taking a closer look at children at work in the classroom, and tape-recorded conversations with children about their learning. In one case I observed a group working with calculators to investigate decimal fractions.
I had written some fractions on the board, and they were discovering their decimal equivalents by using the 'divide' key. I used the dictaphone to record my one-to-one discussion with children in the group, and I observed what happened and wrote notes that evening.
I happened to be sitting next to Peter and he had a go at 12, 24, 34, 58, 78 etc. Peter impressed me because he had grouped together all the fractions that made 0.5; all the ones that made 0.75 and so on.
Later, I gave some different fractions: 110, 210, 310, 410, on to 1110. Peter began on these, and was muttering to himself, "this is really interesting". Then he exclaimed, "the numbers are the same". I asked him what he meant, and he showed me that with 210 you got 0.2 and with 310 you got 0.3 etc. He was really pleased with himself.
I asked him to tell me what he thought 910 would be, and he thought hard, paused and eventually said "nine. . . nought point nine." He was right, and was again looking very pleased with himself.
I then asked him to guess what 1010 would be. He said "ten. . . nought point ten . . . nought point one". I showed him that he had already had 0.1 with 110 so he settled for "nought point ten".
When he tried it and got 1, he was momentarily surprised. "Why's that happened?" I asked him. He immediately reached for the coloured rods, and picked up a "ten" rod and ten "one" rods. He pushed them together and showed me how ten tenths made one whole. I then asked him what 1110 would be, and he guessed "nought point 11". The calculator produced 1.1 and, quick as a flash, he reached for the rods, counted out 11 and pushed them against the whole rod.
He then did something interesting. As he was explaining why it made 1. 1, he said "you take away the ten tenths and that leaves one whole and one tenth". He'd muttered something under his breath about "taking away" before, but I hadn't picked up on it.
I asked him to show the rest of the group, and he again repeated this thing about "taking away". The others looked a little confused, and Alan even said that you're "not supposed to be taking away". Peter though insisted that the little line in the fraction (the one that separates the numerator from the denominator) meant you had to take away. I didn't press the matter and neither did the others.
I was impressed by two things in particular: that Peter re-wrote his answers in his book in order to get a better handle on the interesting patterns there, and that he used the coloured rods to explain to himself and to me and the rest of the group what he was doing.
Even though he didn't fully understand, he was able to make it clear to himself with the help of the practical apparatus. Both these things I feel sure come from previous work I have done with Peter and the whole class, partly in emphasising the importance of methodical approaches to problem solving and partly the use of practical apparatus like rods. He was visibly transferring skills to this new situation.
He had a profound confusion about the formal notation of the fraction, taking it to mean a subtract sign. At first sight this is a stupid error, but in fact, I think it is perfectly reasonable. In order to illustrate his point about 11 tenths being equivalent to one whole and one tenth, he had to push his 11 tenths up against a "whole" rod and a tenth rod, and then take away the tenths to show left behind the one whole and one tenth. It is a rational error, and it arises from his creative use of concrete apparatus to make sense for himself what is happening.
The other interesting thing, of course, is that I did not come to this conclusion about the reasonableness of Peter's error until long after the activity was over, and I was sitting at home listening to the tape and jotting down notes about it. It was through observing the activity a second time, that I was able to come to this (more sophisticated) understanding of the learning that was taking place.
I suspect the whole business of trying to fix levelled statements to young children's learning is a complete waste of time. I cannot see any benefit to me as a teacher, to the children or to their parents, of being able to say, "this child is on level 3". The statements of attainment or level descriptions cannot possibly do justice to the richness, complexity and subtlety of young children's learning.
I feel we've gone horribly wrong in our thinking about assessing this learning. What worries me isn't just that our assessments might be wrong, inadequate and unhelpful (though I think they often are). What really scares me is that the new orthodoxy on how we assess learning is infecting and damaging the way we teach. Maybe we are in danger of forgetting how to listen to children, to talk with children, and to watch children learning.
In the clipboard culture of the modern classroom, if we are not careful we will be reading children's stories to count the full stops and capital letters but we will forget how to respond to their writing as a piece of literature. We will observe their scientific experiments, and dutifully record their predictions and explanations, but find ourselves unable to respond when the experiment or the conversation veers off in an unexpected direction.
We will be so concerned about ascribing a level to a child's achievements, that we won't notice the complexity and the subtlety of what children such as Peter are doing in every primary classroom in this country.
Maybe I'm wrong, I certainly hope so. What I think we desperately need, though, is a proper discussion in the profession and beyond, about why we want to assess children's learning, what we are trying to find out, and about the proper relationship between teaching, learning and assessing.
Peter Strauss teaches at Greythorn primary school, Nottingham. All the children's names have been changed.