Sunday, 19 April 2009

Remarks on the Objectivity of Assessment

In far distant student days, I recall one day when I entered the main sitting room of the flat where I lived. The fire was full on and the room seemed intensely hot. I looked at the thermometer, and the temperature read twenty-three degrees Celsius.
Since I thought this excessively hot and wasteful of heating, I asked the two other people in the room to turn the heating down, if not off altogether. I was given the reply that they were quite cold - "freezing" said one - and even through I tried to explain the temperature reading to them, they remained adamant that it was cold.
Now "warmth" and "cold" are, by and large, subjective measurements of temperature - they depend upon your bodily sensations. A simple experiment will demonstrate the validity of this. If you are in a warm atmosphere then any reduction of heat, however slight, is felt of as being "cold"; but if you come in from a bitterly chilling winter day to a moderately heated room (say from ten degrees Celsius to fifteen), then the room will feel "warm". Yet the room is identical in both cases! As to whether the fire stays on (and burns up costly fuel) depends very much upon who has control of the situation - who has authority.
On the other hand, a thermometer can be used as an independent and impartial arbiter. Yet even this "objective measure" calls for a degree of co-operation. Both sides must agree that, no matter what their sensations tell them, twenty degrees (say) is quite hot enough.
Now it seems that in the sciences - in particular, physics, chemistry or mathematics - an independent arbiter can be agreed upon as a matter of  co-operation, rather than being imposed by authority. But even if the means of checking results is forced upon the children by the teacher, it is still true that the appeal is to this as arbiter, and not the teacher; the means of checking are objective, even if not a matter of choice. In such a situation, the teacher is not an "authority" but may even be found to be mistaken. It might even be discovered that, as rumoured, mathematics teachers are notoriously bad at addition. But when one moves into the realms of English, History and French, there are less well-defined objective standards.
French has an objective standard, but part of this rests upon "educational French" which rules colloquial French as inadmissible. For instance, even educated French people don't fuss about the different usages of "savoir" and "connaitre"; this is more an artifice of grammarians (who have not come to terms with the scientific investigations of Chomsky) than an "eternal truth".
And in the English language, as we approach such matters as the writing of essays, we move progressively away from any appeal to an independent and objective measure; instead, we enter the domain of "the authority". An assessment of an essay could use rigorous objectively laid down criteria - an example might be to mark according to the vocabulary range of the work. But the problem with any method like this is that it is too mechanical - a computer could quite easily be programmed to carry out the marking. The trouble is that it cannot accurately measure the diversity of approach - between, for instance, an argumentative essay and one that tells a tale. To be more precise, this calibrates only the form of the essay and not its content.
On the other hand, an assessment of content must be largely subjective and, therefore, dependent upon the goodwill and health of the teacher. For instance, if you were the teacher, you might mark an essay entitled "Remarks on the Objectivity of Assessment" by making various objective calibrations on the form - such as measuring spelling mistakes, typing errors and readability (although how to differentiate between the first two is problematic!). But instead of this approach, you might prefer to award marks for the imaginative and creative thought involved in the writing, of which there is probably little. Yet another approach might be to consider the logic and so assess the arguments and analogies used. Generally, I imagine that any examiner would mark it on a combination of these methods, with no clear single strategy.
What do I conclude from this? As I have argued, objective marking in a subject such as English will depend upon the form of the material (words used, grammatical construction, spelling etc. ) while subjective marking depends upon the opinion and perspective of the marker. In the latter case, it is the opinion of the examiner that counts, and while that may claim to be well-informed, it is hard to see how this approach can specify a mark with any degree of objective validity.
I do not think this is a particular bad state of affairs in so far as it is recognised. However, when it is ignored, we undermine confidence in all attempts, however feeble, to assess written material. We yield to the totalitarian domain of the "expert" who alone has sole authority to judge rightly. But when the subjectivity inherent in marking is recognised and the examiner is very much aware of self-limitations and fallibility, then, on this basis, I think that it can be justified.

No comments: