"Making Good Progress?" Chapter 4 Analysis and Review

This is part 4 of my review of "Making good progress?" by Daisy Christodoulou. You can find the review index and my analysis of chapter 1 HERE.

Chapter 4: Descriptor-based assessment
 
Chapter 4 begins with an overview of descriptor based assessment. This is by far and away the most common form of assessment used in schools today, in both its formative and summative uses. Christodoulou notes how the summative descriptors for Key Stage 3 rapidly expanded to become APP criteria, designed for formative use. She also notes how many schools’ post-levels solutions are also based on the generation of generic, linear descriptors. This is a topic I find quite interesting and have written on the subject HERE, HERE and HERE.
 
Christodoulou goes on to explore the uses and limitations of descriptors as a tool for formative assessment. She argues that descriptors do not allow teachers to analyse the performance of students, or distinguish between “fleeting performance and genuine long-term learning” (p.85). Whilst I do take her point here, certainly in terms of KS3 level descriptors or GCSE bands, there seems to be a conflation of descriptors and generic descriptors going on here which is not fully acknowledged.  I would argue that it is possible to create useful performance descriptors for individual assessments (an essay for example), and to tailor these for the purpose of assessing both summatively and formatively in that particular task. Indeed, Burnham and Brown have written on this theme in Teaching History 157. It is also possible, I believe, to use descriptors of gold-standard performance to tailor a descriptor-based mark scheme for formative purposes. In a sense, it is about whether the descriptors we use have been properly adapted for the purpose we want to use them for, as well as whether or not they were valid descriptors in the first instance. In the case of APP grids, I would say they answer to both of these questions is in the negative.
​
Christodoulou then uses an example of how summative descriptors cannot help teachers to analyse the performance of their students. Her example of the need for vocabulary to make inferences is a good one and shows the limits of her example mark scheme in diagnosing the formative needs of the child. However, this also assumes that the teacher is only using the summative mark scheme to formatively assess the child in question. I am sure that some teachers only use the mark schemes to form their views on pupils’ progress, however I know that our teacher training course at Leeds Trinity tells trainees that this is a very poor method of assessing formatively. So once again, Christodoulou seems confuse the existence of a particular form of assessment with evidence of how teachers practice assessment on a day to day basis. What is missing here, indeed what this book is crying out for, is a detailed study of what teachers do in the classroom. Again, I suspect that the use of descriptors is more connected with a practical concern (driven by policy) to get the greatest possible advantage in public examinations.
 
At this point I became quite interested in what happens in Ark schools, where Christodoulou serves as head of assessment. Trawling through the websites of all Ark secondaries, I managed to find 2 schools adopting Christodoulou’s suggestions and a large number still using KS3 levels.  I did not get around to investigating Key Stage 4, but the use of summative, generic descriptors certainly seemed to still be in place in 4 schools, and APP grids in another. Most did not publish any information about assessment. This may not mean very much in the long run, but it is an interesting study in power dynamics and the ways in which schools respond to changing external demands. 

In the next part of the chapter, Christodoulou suggests that it might be more valid to test spelling and vocab separately from the final tasks.  She glosses over the inherent sampling issue of this however and does not go on to discuss the phenomenon where students can accurately recall information but struggle to apply in in context – the very aspect which APP grids were meant to solve. I am in no way defending APP as an approach here, I tend to agree with Christodoulou that it encouraged lazy assessment based on unhelpful criteria. However, I also think this misuse was driven by that same misuse of power and accountability which I have discussed before, rather than genuine beliefs among teachers that the APP grid was anything more than hoop jumping.
 
The remainder of the chapter deals with the issue of generic feedback, which I feel was covered suitably in Chapters 1 and 2. Here it feels that we are labouring the point by using extreme examples of practice. In the history example for instance, Christodoulou gives an example of using generic descriptors to offer feedback on an essay on the Battle of Hastings. However, I would contend that most teachers would check the knowledge of the essay as well, thereby overcoming much of the damage of the generic mark scheme. Even this might be overcome by following Burnham and Brown’s suggestions referred to above.
 
In the further example given of a history exam on Stalin, Christodoulou suggests that such a question might quickly highlight misunderstandings. But this question actually opens up a continuum of answers for which multiple choice is probably not appropriate, as a case might be made for options A and C and a series of other potential options are missing, narrowing the scope of the potential answer. Would I use this as a short formative task, yes. Would I want it to be the only assessment of this, no!

​The final section of the chapter goes back to the issue of the validity of descriptors for summative assessment and raises a number of useful and relevant points on their limitations. I found this uncontroversial and fairly measured. However, in subjects like history it is very difficult to depart from descriptors without affecting the validity of that assessment for determining the historical understanding of the student. All written subject will need at least some qualitative aspect, so I am interested to see how Christodoulou deals with this in her next chapter.