Daisy Christodoulou’s ‘Making good progress?’ My initial thoughts

I have just finished reading the excellent ‘Making good progress?’ and my head is brimming with ideas. What I particularly like about Christodoulou’s writing is that it is clear, concise and gets right to the point. So often, academic writing can be dense and rather turgid – getting through it can be a badge of perseverance. Not so with ‘Making good progress?’ which I easily read in one day. That’s not to say it makes for light reading, but simply that there is not a single superfluous paragraph in it. And there is so much to think about – I’m sure I will be delving in and out of this book regularly for the foreseeable future, as I develop as a teacher.

The book has implications for everyone in teaching, from the humble classroom teacher to subject leaders, head teachers and examination boards. From my perspective as a trainee history teacher, the focus is less on summative assessment (over which I will have little control) and more on formative assessment. What I propose to do in this blog is to summarise my main takeaways from the book and then suggest some ways in which this knowledge can be applied in my practice this year. These will be general ideas to start with, which I hope to develop further in future blogs.

My main takeaways from the book

There are broadly two approaches to teaching skills: teaching them directly (the generic-skill method) and teaching them indirectly (the deliberate-practice method). The generic-skill approach involves teaching the desired skill directly, with the activities mirroring the desired end product. So for example, if we wanted to teach students how to write a history GCSE essay, then the activity would be to write such an essay. This is the main type of teaching I have observed during my year as a learning support assistant at my last school. A particular topic is studied and then students attempt to answer a GCSE style question on it.

The deliberate-practice approach, on the other hand, does not involve practising a skill in its final form. Instead, the skill is broken down into different component parts and these are practised until mastery is reached. Only then, does the student attempt the actual intended skill to be learned, such as the GCSE essay mentioned above. Formative assessments will differ according to the approach that is taken. In a generic-skill approach, formative assessments will look very similar to the summative assessments, so they will most likely be past GCSE questions. However, formative assessments in the deliberate-practice approach would assess those different component parts identified above. In Christodoulou ‘s words:

The main problem with the generic-skills approach is that skills are not generic, they are specific. Thus it is nigh on impossible to teach discrete skills such as problem-solving or critical thinking. You cannot get better at problem-solving simply by practising solving problems. This is most clearly exemplified in experiments carried out in the 1970s by Herbert Simon on chess players. Simon tested expert, intermediate and novice players by showing them for 2-10 seconds a chess board taken from the middle of a real game and then asking them to reproduce it from memory. In this test, the expert chess players were able to reproduce on average two thirds of the pieces correctly while novices could only remember 20% of the pieces.

When the experiment was repeated, this time with random pieces on the board rather than copying a position from an actual game, the novices and experts performed equally poorly. The implication of this experiment is that skill is not easily transferred, even in similar domains and that is because it is domain-specific. From his experiments with chess players, Simon hypothesised that expert players have between 10,000 and 100,000 chunks of chess positions stored in long term memory. So it was easier for them to remember the chess pieces in the first experiment because they corresponded to chess positions already logged in their long term memory. When random pieces were displayed in the second experiment, they were unable to remember them any better than the novices because there was no identifiable chess position and they had to rely, not on information stored in their long term memory but on what they could remember using their working memory.

According to Christodoulou, experts in all fields

‘depend on rich and detailed structures of knowledge stored in their long term memory. These structures – often called schema or mental models – are what allow the expert to encounter new problems and solve them with such ease.’

In other words, when experts are problem solving, they are not exercising some generic problem solving muscles. What they are actually doing is using a store of specific mental models in their long term memory to help solve the new problem, as with the expert chess player example. If experts are called upon to solve a problem using only working memory, they become no more capable of problem solving than the novice. Working memory is limited to as little as three or four items and relying solely on it is not particularly effective.

‘We need the help provided by the mental models stored in long-term memory and in order to get that help we need to acquire such mental models in the first place.’

Thus, acquiring mental models and storing them in long term memory is the necessary precursor to being able to perform particular and specific skills. This is what is behind the deliberate-practice approach. By practising each individual component parts, the pupils are building up those vital mental models that are needed to perform the skill being learned. In the generic-skill approach, however, there is not much opportunity for building up the required mental models. Because this approach involves simulating the end skill, for example writing an essay, working memory is taken up by the effort of writing the essay and there is little opportunity for building up new mental models. The act of performing a skill affords fewer opportunities for learning that skill than deliberate practice of the component parts.

‘Paradoxically, methods of teaching which ask pupils to do real and complex tasks will prevent pupils from developing the mental models they need to actually be able to solve those real and complex tasks.’

The implications of this are, to me, mind blowing. Practising essay writing, particularly at the novice stage, does not really help improve essay writing. Deconstructing the knowledge required to write a good essay on a specific topic, and practising the component parts, will yield far better results. This could turn our teaching of history on its head. With this new insight, I plan to take a closer look at the schemes of work for the year 7 and year 8 classes I will teach next term. If, for example, one of my objectives will be to get pupils to write about why William won the battle of Hastings, then I will have to think carefully about all the component parts of knowledge they will need to achieve this task, and then teach and assess those component parts before attempting the final written piece of work.

‘If we want pupils to develop a certain skill, we have to break that skill down into its component parts and help pupils to acquire the underlying mental model. Similarly, when developing assessments for formative purposes we need to break down the skills and tasks that feature in summative assessments into tasks that will give us valid feedback about how pupils are progressing towards that end goal.’

So what is the best way for pupils to acquire these mental models and for us to assess their progress? Here, Christodoulou discusses the various methods that are currently used to assess pupil progress and critiques both the descriptor-based and exam-based models. Descriptor-based comments are ‘effectively just grades in prose’, they suffer from a lack of reliability and are prone to teacher bias (which statistically has been shown to be biased against disadvantaged pupils).

Exam-based assessments also have problems, the main one being that they don’t allow teachers to infer and diagnose exactly where the pupil is misunderstanding or has a gap in knowledge. This is particularly the case in subjects such as English or History where exam-based assessment still relies heavily on descriptors. The other problem is that tests are samples from a domain – to test the entire domain, tests would have to be extremely long. Thus they are not entirely able to give the teacher information about how well pupils are understanding all the required component parts of the skill being learned, just a sample of them. The more difficult a question is in an exam, the more component parts are required to be able to answer it. If the pupil gets this difficult question wrong, it is not always possible, through logging on a spreadsheet, for it to be apparent which particular component let the pupil down and therefore where in particular the teacher needs to reinforce knowledge.

Having described the problems with these types of assessment, Christodoulou then goes on to suggest an alternative way.

‘A good assessment system must not only clarify the current state and the goal state, which it can do through the use of summative assessments, but it must also establish a path between the two: the model of progression. An assessment system has to make clear the link between the activities being done in the individual lessons, the summative exams that come at the end, and the wider domain of expertise the exam is sampling from.’

This links up with Michael Fordham’s ideas about progression models, in which the curriculum itself is the progression model. According to Fordham:

‘If a student has learnt the curriculum, they have made progress. In a curriculum, we set out what we think pupils ought to learn. If they learn what we have set out in the curriculum, then they have by definition got better at history.’

The most important aspect of making progress is thus to identify the components of knowledge to include in the curriculum, break them down into chunks and sequence them in a coherent way. Christodoulou is clear here, that what we don’t want to be doing is to use the final exam – say SATs or GCSEs – as the progression model. We do not want to teach to the test, because tests are only samples of a domain.

This leaves us with the task of identifying what exactly we want our pupils to learn and to remember. Going back to my scheme of work for next year, I need to sit down and think about what specifically I want my pupils to learn and remember. In addition to historical knowledge, I will want to think about specific words and phrases that I want my pupils to learn. Leading researcher Isobel Beck recommends:

‘teaching about 400 words a year for the first ten years of education. Specifically she recommends teaching a set of words which she calls “Tier Two” words… Tier Two words are common in prose but are less common in everyday speech: words such as “eradicate”, “inevitable”, “restrict”, “variation” and “industrious”. We cannot rely on pupils picking them up from everyday interactions, but if pupils are to become fluent readers of sophisticated texts, they are vital.’

In the process of thinking about what exactly I want my pupils to learn next year, I will need to come up with a list of Tier Two words useful for historical writing and relevant to the content being taught. These words, along with the specific curriculum components identified, will need to be taught explicitly and practised, with a particular focus on spaced, retrieval practice, which I have talked about before. Interestingly, Christodoulou champions the use of multiple choice tests, something I had not thought to use in the context of history teaching. I will look further into these in my next set of blogs, in which I hope to focus on the specifics of what I will need to teach next year, in light of all the insights I have acquired through reading Christodoulou’s excellent book. For now, I think I have written enough.

Please follow and like us: