The relationship between assessment and correction
It is, of course, perfectly possible to give assessment without correcting, as when a final percentage mark on an exam is made known to a learner without the exam itself being returned or commented on. The other way round is very much less feasible: it is virtually impossible to comment on what is right or wrong in what a learner has done without conveying some kind of assessment. If a correction is supplied, the learner is very aware that this means the teacher thinks something was wrong; if comment is given on why something was appropriate, there is necessarily an underlying message of commendation.
Teachers are sometimes urged to be ‘non-judgemental’ when giving feedback. Although any meaningful feedback is going to involve some kind of judgement It is more useful, perhaps, to accept that there is judgement involved, but to try to make the attitude to this more positive: that mistakes are a natural and useful part of language learning; that when the teacher gives feedback on them, the purpose is to help and promote learning; and that 'getting it wrong' is not ‘bad’, but rather a way into 'getting it ‘right’.
Approaches to the giving of feedback
Below you will find expressions of selected opinions on the nature and functions of assessment and mistake correction; these are based on different theories of language learning or methodologies
Assessment: different opinions
Audio-lingualism
Negative assessment is to be avoided as far as possible since it functions as 'punishment' and may inhibit or discourage learning. Positive assessment provides reinforcement of correct responses, and promotes learning.
Humanistic methodologies
A crucial function of the giving of assessment is to preserve and promote a positive self-image of the learner as a person and language learner. Assessment therefore should be positive or non-judgemental.
Skill theory
For successful acquisition of a skill, the learner needs feedback on how well he or she is doing; hence the importance of the provision of constant and honest assessment
The correction of mistakes: different opinions
Audio-lingualism
Learner mistakes are, in principle, avoided by the limiting of progress to very small, controlled steps: hence there should be little need for correction. The latter is, in any case, not useful for learning; people learn by getting things right in the first place and having their performance reinforced.
Cognitive code-learning
Mistakes are regrettable, but an unavoidable part of learning; they should be corrected whenever they occur to prevent them occurring again.
Intel-language
Mistakes are not regrettable, but an integral and important part of language learning;
correcting them is a way of bringing the learner's ‘intwerlanguage’ closer to the target language.
Communicative approach
Not all mistakes need to be corrected; the main aim of language learning is to receive and convey meaningful messages, and correction should be focused on mistakes that interfere with this aim, not on inaccuracies of usage.
Monitor theory
Correction does not contribute to real acquisition of the language, but only to the learner's conscious 'monitoring' of speech or writing. Hence the main activity of the teacher should be to provide comprehensible input from which the learner can acquire language, not to correct.
Assessment
Most of the feedback we give our learners is ongoing correction and assessment directed at specific bits of learner-produced language with the aim of bringing about improvement; the type of evaluation involved here is sometimes called 'formative', since its main purpose is to ‘form’: to enhance, not conclude, a process. Distinct from this is the evaluation usually termed ‘summative’, when the teacher evaluates an overall aspect of the learner's knowledge in order to summarize the situation: how proficient he or she is at a certain point in time, for example, or how much he or she has progressed during a particular course. Summative evaluation may contribute little or nothing to the ongoing
Teaching/learning process; but it is a part of the teacher's job, something we need to know how to do effectively.
Below are descriptions of various ways of gathering the information which will serve as a basis for assessment, and of some common criteria used for assessing it.
Gathering information (1): Tests
The most common way of gathering information for assessment is through tests; the usual criterion is an arbitrary level which the learner is expected to have reached; and the result is generally expressed through percentages.
Gathering information (2): Other sources
There, are, however, various problems with tests as a basis for summative evaluation: they are a one-off event which may not necessarily give a fair sample of the learner's overall proficiency; they are not always valid (actually testing what they say they are) or reliable (giving consistent results); and if they are seen as the sole basis for a crucial evaluation in the learner's career, they can be extremely stressful.
Other options do, however, exist. These are summarized below.
1. Teacher's assessment. The teacher gives a subjective estimate of the learner's overall performance.
2. Continuous assessment. The final grade is some kind of combination of the grades the learner received for various assignments during the course.
3. Self-assessment. The learners themselves evaluate their own performance, using clear criteria and weighting systems agreed on beforehand.
4. Portfolio. The learner gathers a collection of assignments and projects done over a long period into a file; and this portfolio provides the basis for evaluation.
Criteria
Having collected the 'evidence' of the learners' proficiency in one or more of the ways described above, the teacher has to decide how good it is? The following are some of the possibilities.
1. Criterion-referenced: how well the learner is performing relative to a fixed criterion, where this is based on an estimation of what it is reasonable or desirable to demand from learners at the relevant point in their development (age, career, level, stage of a course).
2. Norm-referenced: how well the learner is performing relative to the group. In this case, a group of slow learners would be assessed according to different, easier, norms than a group of faster ones.
3. Individual-referenced: how well the learner is performing relative to his or her own previous performance, or relative to an estimate of his or her individual ability.
What criteria do/would you yourself use in assessing learners' performance? Would you combine different criteria? Would you take into account learners' effort, motivation and progress in deciding on a final grade?
Correcting mistakes in oral work
There are some situations where we might prefer not to correct a learner's mistake: in fluency work, for example, when the learner is in mid-speech, and tocorrect would disturb and discourage more than help. But there are other situations when correction is likely to be helpful.
The recommendation not to correct a learner during fluent speech is in principle a valid one, but perhaps an over-simplification. There can be places where to refrain from providing an acceptable form where the speaker is obviously uneasy or 'floundering' can actually be demoralizing, and gentle, supportive intervention can help. Conversely, even where the emphasis is on getting the language right, we may not always correct: in a grammar exercise, for example, if the learner has contributed an interesting or personal piece of information that does not happen to use the target form; also, when they have got most of an item right we may prefer not to draw attention to a relatively trivial mistake.
Oral corrections are usually provided directly by the teacher; but they may also be elicited from the learner who made the mistake in the first place, or by another member of the class. Corrections may or may not include a clarification of why the mistake was made, and may or may not require re-production of the acceptable form by the learner.
As important as what the correction consists of ishow it is expressed: gently or assertively, supportively or as a condemnation, tactfully or rudely. On the whole, of course, we should go for encouraging, tactful correction; but it is less easy to generalize about gently/assertively: some learner populations respond better to the one, some to the other. In general, in fact, learner responses to different expressions of feedback are often surprising: a teacher correction that seems to an observer a humiliating ‘put-down’ may not be perceived as such by the learner to whom it was addressed; or an apparently gentle, tactful one may give offence. A good deal of teacher sensitivity is needed here.
Tests and Testing
People vary very widely in their reactions to tests. Some like the sense of challenge; others find it unpleasant. Some perform at their best under test conditions, others perform badly.
Thus, it would be a mistake to come out with sweeping statements like: 'People get very stressed when they are tested', or 'Tests are unpopular'. The amount of unpleasant stress associated with a test depends on various factors, at least some of which may be under the control of the teacher: how well the learners are prepared for it and how confident they feel of success; what rewards and penalties are associated with success or failure (how important the results are perceived to be); how clear the test items are; how easy the test is as a whole; how often such tests are given; and so on.
Types of tests
Questions and answers
These can be used to test almost anything. The more 'closed' the question is (that is, the fewer the possible options for correct answers), the easier the item will be to mark. It is fairly easy to compose and grade closed-ended questions; more open, thought-provoking ones are more difficult, but may actually test better.
2. True/false
This does not directly test writing or speaking abilities: only listening or reading. It may be used to test aspects of language such as vocabulary, grammar, content of a reading or listening passage. It is fairly easy to design; it is also easy to administer, whether orally or in writing, and to mark.
Multiple-choice
This may be used for the same testing purposes as true/false items; it does test rather more thoroughly since it offers more optional answers and is obviously very easy to mark. It is administered more conveniently through writing; but note that since the reading of the question-and-options is fairly time-consuming, the process of comprehension of the actual question items may take more time and effort than the point ostensibly tested, which raises problems of validity. Another important problem is that good multiple-choice questions are surprisingly difficult to design: they often come out ambiguous, or with no clear right answer, or with their solutions over-obvious. They are to be approached with caution!