This blog post discusses Daisy Christodoulou’s new book about assessment for learning; I hope that my description and reflection below eventually leads to more teachers choosing to read the book as well as a general spreading-of-the-word about what constitutes effective assessment. Firstly, I’ll start by attempting to summarise the book, then I will talk about how primary schools might implement the recommendations and how this could link to Daisy’s exciting vision of a grander scheme.
Here’s that summary I mentioned earlier:
- Ch. 1 AfL hasn’t transformed schools: Daisy describes the importance of knowing the difference between teaching generic skills and teaching knowledge with more specific skills. Formative and summative assessment look pretty similar for the former teaching method (for example, practising problem solving in order to solve a plethora of problems), but very different for the latter (eg. learning and practising short multiplication in order to solve multiplication problems).
- Ch 2 Curriculum aims and teaching methods: Daisy goes into greater detail, backed up with considerable research-based evidence, about the folly of trying to develop general skills via project based learning, even though Ofsted inspectors clearly love this kind of thing. Essentially, the research shows that PBL is limited by working memory and to solve any particular problem, you actually need lots of domain specific knowledge and vocabulary (which needs to be taught and learned) to help you out. Daisy then makes the case for direct instruction with memorisation and practice of the pertinent bits and pieces that eventually can come together to allow someone to perform a more complex, domain-specific task. For this to work, we need to remember that assessment for learning (along the way) needs to assess those bits and pieces, but assessment of learning (at the end) needs to assess the final ability to do something complex (but still quite specific).
- Ch 3 Making valid inferences: Daisy clarifies a few meanings associated with testing such as reliability and validity and then goes on to describe the relative reliability and validity of two different types of tests: tests that judge ‘quality’ (like an essay) and tests that judge ‘difficulty’ (like a maths test where questions get progressively harder). Can you integrate formative and summative assessment effectively? No, not really because assessment is more or less reliable or valid depending on what it’s being used for. Primary teachers/leaders of English might want to check out Engelman’s ‘Expressive Writing’ as an example of a teaching programme that has very effective uses of formative assessment, with the possibility that summative assessment could also be integrated quite easily.
- Ch 4. Descriptor based assessment: Think of the nightmare of APP! Yet, descriptor based assessment, like some kind of shapeshifter, hangs around in various guises and some schools even use multiple versions all at once. The average primary teacher also has the additional *joy* of having to assess ALL the subjects, not just one! For formative assessment these tests are too specific, only describe performance and can’t be used to inform improvement. For summative assessment judgements they are inconsistent, difficult and prone to bias; descriptors too can be interpreted in too many different ways. So, all round they’re not good. What can we use then?
- Ch 5. Exam-based assessment: Although designed for summative assessment, we can use exam question analysis for formative assessment. Surely that’s all sorted then? Not so fast! Test don’t directly measure those little bits and pieces of learning that are needed in order to do something complex, so it’s hard to identify areas of strength and weakness; further, how can a simple grade tell us what we need to do next? (From my perspective, at least a low grade gives parents a reason to take away privileges until the child pulls his socks up; sometimes what is needed is not specific revision or new learning via the tired teacher, but putting in more effort in general!) However, exams are pretty good for summative purposes because they are reliable and consistent, although you have to watch out that lessons don’t end up being about teaching to the test.
- Ch 6. Creating a [really good] model of progression: I was absolutely not expecting The Textbook to turn up in this book, but turn up it did! On any measure, the textbook is a far superior (communication) model of progression than an APP style grid and if you don’t believe me, you’ll simply have to buy Daisy’s book yourself! Formative assessment then needs to be matched to the different elements that are being assessed.
- Ch 7. Improving formative assessments: Formative assessments need to be specific, frequent, repetitive and recorded as raw marks. Daisy recommends multiple choice questions partly because you can also rope a bit of technology in to help lighten the marking burden as well as provide further insight on whole-class gaps in understanding and knowledge. Further, you can deliberately pop in answers that smoke out typical misconceptions, as well as choose to test what has been taught in previous units to see if what has been taught has actually stuck around long term! Research also shows that frequent testing is also good for long term memory (because you are having to repeatedly think about certain learning points). Swift intervention is also a key feature of good systems of formative assessment, daily if possible.
- Ch 8. Improving summative assessments: Rubrics are pretty tricky to work with and lead to all sorts of weird consequences in the classroom. Daisy’s solution is to use comparative judgement for summative assessments that are in essay/story form. As I understand it, with comparative judgement you compare each essay to annotated examples in order to arrive at a final mark or grade (I need to find out more about this). Further, standard conditions as well as scaled scores are also key features of a good summative assessment system.
- Ch 9. Integrated assessment system: Daisy’s recommendation is a textbook to bring together a model of progression, with a bank of formative assessments matched to each lesson or chapter. Technology can be really useful here because the assessments can be done online, with the pupil being directed to further study where needed. For summative assessment, you would add, with the help of technology, a yearly ‘difficulty model’ or ‘quality model’ summative task. The former could be taken online with intuitive feedback directing pupils to appropriate questions, and the latter could be linked to a comparative judgement engine. Finally, you would add standardised tests that are not linked to the curriculum, but would give you additional information about the child’s reading age, for example.
- Conclusion: Daisy reminds us that all these components, and the technology to support such an integrated system, already exist. What is needed for it all to work is for organisations to come together to make this happen, ideally, on a national scale. Improved measurement can be ‘transformative’ just as it has been for other fields such as medicine and economics.
This was a lot to get my head around! Half way through the book I was thinking that what is required of primary teachers to improve assessment is pretty much impossible. I already struggle as it is with our various systems of assessment, coupled to spreadsheets flying about all over the place, targets, meetings and the general unease with the incontrovertible fact that, unlike quantum physics, measuring something like progress or achievement doesn’t actually change anything! This is because there is hardly any room, save for a few tweaks here and there, to change teaching for children who flag themselves up quite frequently as needing serious intervention. Further, if you think about the maths, with each additional assessment you could end up, in theory, with the ‘solution’ that each and every child in your class needs private tuition and I don’t know about you, but I don’t have a teacher-cloning machine, yet.
Luckily, I stuck with it to the end and ended up not just heaving a sigh of relief, but actually feeling very positive about the whole situation. In fact, I had a few ‘lightbulb’ moments!
So, if we run with Daisy’s proposal of a joined up system involving a textbook, linked and computerised formative and summative assessments combined with assessments of general skills and ability, how could this work in a primary school?
For me, I’m thinking about doing whatever I can with the limited power I have. This would probably mean regularly creating question banks online for children to do formative assessments in the classroom. With a little more power, say, leading mathematics education in a school, I would probably look to investing in a good available textbook such as The Shanghai Maths Project currently in progress with Collins, implementing teaching methods that are used in Shanghai which emphasise those crucial components of knowing maths facts and algorithms off by heart, as well as creating online formative assessments and summative assessments linked to the textbook. This would cut out the need to create a progression model because the good people over at Shanghai have already done it (for God knows how many years now), including factoring in all those common misconceptions in a proactive way, which is much better than the current situation whereby English primary maths teachers tend to encounter misconceptions (if they spot them) in a reactive way. With the kind of influence SLT have I would also look to manage HR and lesson logistics such that daily interventions could take place, as well as instigate the kind of whole-school discipline that leads to children acquiring a more scholarly disposition and respect for the teacher, putting the (always interesting) subject firmly at the centre of learning. If there were the opportunity to work with a group of primary schools using the same textbook and assessments, I’d be looking to amass that data generated by online formative assessments and do some serious data analysis, in addition to creating or using standardised end-of-year assessments with plain-English reporting to parents. As Daisy said, there’s no reason why we can’t utilise a bit of ‘gameification’ in formative assessment which would help keep children motivated (I talk a lot about ‘PBs’ in my current class) as well as give them the means to describe their progress to their parents.
But what about other subjects? The lone primary teacher cannot possibly do this for all the subjects and besides, the DfE haven’t actually specified what should be on, for example, the humanities curriculum on a year-by-year basis. This is also one of the reasons why most primary schools cling to a skills-based, child-centred curriculum, with secondary teachers finding that year 7 (particularly the disadvantaged children) don’t seem to know the countries in Europe even though they will have ‘discussed’ (fantasy) maps, or how to spell photosynthesis even though they will have ‘collaborated on an investigation’ of plant growth. For change to happen in order to massively improve assessment, and this would help alleviate a big part of the class teacher’s workload, collaboration on a national scale is the order of the day.
Who’s with me?