Why do GCSE and A-Level exams get it so horribly wrong?

As GCSE and A-Level reforms come into reality over the next year, I am left reflecting on the latest stage in a story that has been developing in England for over two decades: the way in which sources are used in history exams. I have written before about why sources matter in school history, and in a recent post I raised the suggestion of ‘anthologies’ and ‘set texts’ as a means of overcoming some of the problems involved in including sources in history exams. This is just a quick post in an attempt to give some more air time to what is currently one of the biggest problems in terms of how history is assessed, not just in England, but internationally.

Let’s take a look at an example. This is an example question from one exam board, though most of what I say here applies to all of our exam boards and I should be clear here that I am not setting out to demonise any one board, or even the boards in general: I have written before about the difficulties the boards face. This is a question from the new GCSE History sample assessment material.

One of the criticisms that has been made at length about the use of sources in history lessons in schools has been the tendency to use very short extracts that are not seen in the context of the wider document. Source A is a classic example of this: a short snippet from a speech. We are told nothing in the exam about what the rest of the speech said, nor about what the context was in which that speech was made. If a pupil knows something about Churchill and what he did as First Lord of the Admiralty, then we might be able to guess at the wider context, but given that this is a paper on Conflict and Tension in Europe between 1894 and 1918, it would probably be too much to assume that a pupil had been taught specifically about the early career of Churchill and what exactly it was that he got up to as First Lord of the Admiralty. Pupils are asked here to use their own contextual knowledge, but in practice this is likely to be based more on the wider geopolitical situation in the early twentieth century than specifics related to this source.

So, we start off by giving pupils very little to go on: a short snippet from a longer speech, and a few bits of contextual information. What are pupils then asked to do with this source? Well, the question demands that they decide the strengths and weaknesses of the source for understanding Anglo-German rivalry. The idea here, of course, is that the exam tests a wider domain of knowledge and skills that constitutes the practice of the academic historian. Historians spend their time looking at archival records and using their study of these sources to construct accounts of the past. The pupil in this question is being asked to make a judgement about what the strengths and weaknesses of Source A for historians who want to understand Anglo-German naval rivalry before the First World War.

But here is the problem: you will not find a single historian of this period who even begins to try and decide how helpful a source is who has not (a) read the whole speech (or at least a good chunk of it) and (b) quite extensive knowledge of the context in which the source was produced. I hate to generalise about the practice of historians, but I am pretty confident that the process of archival research bears almost no resemblance to what pupils are asked to do in the exam. In speaking recently with another history teacher (who marks for a different board to the one shown here) he said he was shocked at the crassness of even the best answers given in GCSE History exams. In an attempt to make the GCSE History exam in some way ‘authentic’ to the practice of history, all the exams do is end up getting pupils to perform a task that is pretty far removed from what it is that historians actually do.

And yet I understand why all of our exam boards set questions like this. The DfE regulations stipulate that pupils must be assessed on their handling of sources, and an exam can only be so long: pupils cannot be expected to read a long speech and all of the contextual knowledge needed to make sense of a speech for every question they are going to answer. In squaring this circle, the exam boards give us short extracts and very limited contextual information. This has been the model used for decades.

How do we escape this problem? One option would be to lose sources completely from exams, and I can very much see the argument for this. The vast majority of those who take GCSE History will not go on to study the subject at university and be disabused of their corrupted view of historical practice, and one might argue that it is better for them not to be loaded up with these misconceptions in the first place. I do, however, feel there is still a place in history exams for questions based on sources. To do this, however, we have to solve the problem that pupils need (a) to see source material at length and (b) to know a great deal about the context in which the source was produced.

I am increasingly convinced that the only way in which this is possible is to have the sources on which pupils will be examined set out in the specification, rather like English Literature exams specify the poems on which pupils will be asked questions. I am not suggesting that pupils necessarily need to take a ‘source anthology’ into the exam, but I am suggesting that, when they go into an exam, they should not be encountering sources ‘cold’: there is simply no justification for doing this in terms of the actual disciplinary practice of history.

The boat has been missed for the current round of GCSE History reforms, but the exams will I am sure be reformed again in a few years’ time. When this happens, history teachers need to be ready to show what a quality examination involving sources might look like. As a community we now have a pretty good idea of where things have gone wrong in the past, but we have not yet come up with the solutions that exam boards need in order to move on from the current poverty that is source-based questions in exams. Source anthologies, I would suggest, are an obvious starting point.