Should exam questions be predictable?
I have over the last half term started teaching a Year 11 class for the first time in a couple of years, and I am reminded of just how much work has to be done on teaching pupils exam technique. I do not actually mind teaching exam technique: things like learning how to pace oneself, how to allocate time for marks awarded and how to plan your answers are all useful exam techniques that it is important to teach students.
What I find very frustrating, however, is that I have to teach them a number of ‘question types’ that will come up every single year. We teach OCR GCSE Modern World History and there are always questions such as
- ‘Describe… [4 marks]’
- ‘Explain why… [6 marks]’
- ‘What is the message of… [7 marks]’
and so on. The mark scheme for each question type is very consistent from year to year meaning that it is possible to train pupils to jump through the mark-scheme hoops. I don’t know a single school that does not do this: it would be mad not to. The nature of the exam, which consistently sets the same types of questions from year to year, encourages us to teach in this way. I am not convinced that this is a good thing.
To be fair to OCR, this is not something that is peculiar to them. All exam boards do it (and I assume this is true in other subjects as well). I am not completely sure of the rationale, but I suspect the reason for having common question types every year is that it (a) makes the exam marking more consistent, as examiners are looking for similar types or ‘shapes’ or response and (b) makes schools feel like they have more control in being able to prepare pupils for the exams. I get both of these reasons.
But, as ever in education, it is the perverse consequences that need to be considered.
Take, for example, the numerous consultants who go into schools to look at history curricula (nothing wrong with this – I’ve done it myself!). What worries me is when these consultants go in and say things like ‘you need to do more ‘Explain why…’ questions in Year 7 to prepare them for the exam or ‘why ask that type of question in Year 9 when it doesn’t come up on the exam?’
What we see is teachers spending considerable amounts of time teaching children about how to answer particular question types, and this is time that is not being spent teaching the subject itself. I do not know how many hours teachers spend teaching pupils to spot question types in exams and how to structure answers to those questions to maximise marks, but I’d wager it’s a long time. If children can tell you how to structure a 6-mark ‘explain why’ question, but they’re struggling to define what ‘communism’ is, then I think we have a very aggressive examination tail wagging a beleaguered subject knowledge dog.
So perhaps it’s time to move away from predictable exam questions. All they do is encourage teachers to teach exam technique from as early as possible. If exam questions were less predictable (and I mean everything here from the number of questions through to the wording of the questions) then we would be sending a message to teachers that the best way to prepare pupils for an exam is to teach them the subject well: the current system incentivises teachers to spend far too long teaching exam technique. This will not be a popular suggestion either for the exam boards or indeed for teachers who like having as much control as possible over what pupils write in an exam: I would suggest, however, that we need to sacrifice this micro-management for the good of the subjects and pupils that we teach.
Further Reading
http://heatherfblog.wordpress.com/2015/09/19/is-reliability-becoming-the-enemy-of-validity/
Clio et cetera 
Reblogged this on The Echo Chamber.
At a meeting of history HoDs last year one said she spent 50% of teaching time on technique. Another then pointed out that the skills gained didn’t help students at A level as frequently they have to unlearn techniques and grasp new ones. I like the Edexcel IGCSE because there is a significant reduction in question types to drill. You have more knowledge of exam question creation than I do but it seems that the needless proliferation of question types is to satisfy assessment objectives that assume different skills can and are being assessed.
There is also the question of access for weaker students when question are less formulaic. Anyway
I entirely agree that currently far too much time has to be spent on drill at the expense of developing a knowledge of history and things could be better.
The law of unintended consequences: If you moved away from ‘predictable exam questions’, might teachers not end up spending *more* not less time on exam technique rather than subject knowledge? Teachers would feel that they would have to prepare students for almost every question stem / command word possible. This would take far longer than is currently spent.
‘Predictable exam questions’ mean that you can get the exam technique sorted *quickly* so that you can spend more time on the knowledge and understanding.
I agree with your diagnosis of the problem; but disagree with the solution. The system is imperfect as it is, but it would be worse if you abandoned ‘predictable exam questions’.