For years one of the criteria for assessment on Cambridge Speaking exams was accuracy of grammar and lexis at varying degrees depending on the level. Now, accuracy seems to only be welcome in the global achievement descriptor and has been upstaged in the other criteria by “Control”, but what exactly is the difference?
We have just been spending quite a lot of time discussing assessment criteria for Cambridge exams as the annual meetings are taking place at the moment where speaking examiners meet and do tests together to standardize. One of the points that came up in my meeting last Saturday was this control vs. accuracy issue.
Looking at what candidates can do rather than what they can’t do
In recent years we have all started to think more about what candidates can do rather than what they can’t do, but many language teachers find it very hard to shift away from what is, in fact, an extremely normal reaction: to see the errors students make and try to help them stop making them.
In an exam situation though we have to be careful not be in the position of not seeing the wood for the trees. We see the errors and inaccuracies because they stand out, but we don’t see what the candidate can actually produce. This is not a new idea but it is one that is quite hard to get across or even accept. What happens for instance when someone listens to another person? Is is all too easy to miss good language production because it is easy to nderstand and to focus, rather, on the anomalies or inaccuracies.
Some, then, think that control is the same as accuracy but with a different name. In fact, I don’t think it is, and various different degrees of control are mentioned in the criteria so that a candidate would be given one mark for “sufficient control” of the grammar to be able to complete a task and a higher mark for “good control”. This might mean, that in a test like the Key English Test (A2) where candidates are asked, in Part Two, to ask each other questions, an utterance like this would show sufficient control to complete the task:
“Tickets? How many pounds they cost?”
The question form, of course, is not accurate, but there is enough control both of the lexis and the grammar to be able to get the meaning across at an elementary level. What “control” then means to me (I hope someone will correct me if I’m wrong) is a sort of half way house between not being able to communicate and being able to communicate perfectly, a goal not many of us ever reach, in fact. It means looking at the candidates discourse as a work in progress and seeing how much or little he or she can use effectively, rather than identifying a list of errors. At higher levels, of course, more control would be expected but this still means that hopefully examiners will be looking at what candidates can do and not how many errors they are making.
What about higher levels?
Here, of course, is the rub. What counts as control at one level may not be the same at a higher level where the discourse is extended, the message being communicated more complex, and therefore, there is all that much more room for lack of control, but the principle, I think is the same. Small non-impeding inaccuracies should not be over.penalized and more emphasis should be placed on the range of grammatical and lexical resources that are used. After all, I recently saw some examiners getting all hot and bothered about a candidate at a B1 level who had misused “funny”in a written test, and this made such an impression on them that they entirely missed the modal verbs the same candidate had used such as “I won’t be able to…” or “I really had to…”
So, I’m afraid I think the answer is “yes” accuracy is no longer so welcome at the ball, and should at least be accompanied by an awareness of control.