Assessment matters: Assessing skills

Introduction

Most teachers would agree that it's preferable to assess students' abilities in language skills such as reading, writing, listening and speaking and not simply to equate language proficiency with grammatical accuracy. However, methods used to assess such skills are often inadequate at best and in many cases simply do not achieve what they set out to.

What do most teachers do?

I’ll start with one of the productive skills – speaking – as this is one of the easiest to assess.

Most speaking assessments seem to expect students to respond to some kind of stimulus, whether this is some questions, a picture or a posed problem. There’s not too much wrong with such tasks, except they often require the student to understand the written or spoken question – in other words, there is an element of listening or reading involved. As long as some kind of leeway is given for this, and any assessment is based on the production, then there really isn’t much of an issue.

Often, though, one of the criteria is that the student ‘completes the task’ or ‘answers the question’. However, if they didn’t understand the original question then how is this possible? So, there is some element of testing the student’s ability to listen or read. Nothing wrong with that, you might say. And I’d agree, up to a point. The point is that you aren’t just assessing a student's speaking skills but also their ability to comprehend an instruction given verbally or in writing, so you are no longer able to state that it is a speaking (only) assessment.

Similarly with reading, if you expect students to give a written or verbal response, then any assessment is somewhat flawed. However, it is quite easy to assess a person's ability to read. Simple true/false questions or ordering events from a story require no other skill other than reading and understanding and therefore are appropriate.

What about listening?

This is probably the hardest to assess and is almost impossible to test without using one of the other skills. However, it is not totally impossible. TPR (Total Physical Response) is a technique by which a student’s listening ability can be assessed without resorting to speaking, writing or reading. For example, the teacher says, ‘Stand up' and only half the class stand up. In all likelihood the half that didn’t stand, didn’t understand the words. Although we must still be cautious, as it is quite possible that some of the students who stood up did so because they copied the other students. Therefore we need to use other techniques in addition.

So what should we do?

We have two possible solutions. The first is that we stop saying that we are assessing a particular skill, and instead become more realistic and start talking about assessing skills in combination. The other option is that we try and design tasks that focus solely on the skill we are trying to test. Let's take a look at the practicalities of these two choices.

Let’s imagine we give our students a short reading text followed by a set of questions for which they are supposed to write answers. When we assess the task we can then have criteria which includes comprehension of the text (possibly including a vocabulary element) and the ability to write clear answers (possibly including an element of grammatical accuracy and use of own words – as opposed to ‘lifting’ answers directly from the text). We now have a task that is clear and manageable with a realistic set of assessment criteria in terms of the skills involved (reading and writing).

One problem with such an approach is in making the criteria and balance between the skills fair. For example, if someone answers a written question about a reading text, what is the balance between the content and the grammatical accuracy of the answer? This is currently a problem with the way we assess skills. We are still left with a potential problem, where a lack of writing ability can detract from whether the student has understood the reading text or not. A student’s ability, or lack of ability, in one skill may well have a detrimental affect on their overall test score.

The alternative is that we find ways to assess the reading text without resorting to questions that require another skill, i.e. writing the answers or listening to the teacher and responding orally. One way might be to read a text and then put a set of pictures in order. Another would be for students to read a text where the paragraphs are mixed up and to put the paragraphs in the correct order.

Are all the techniques we currently use wrong?

No, not at all. In fact, most of the tasks already mentioned are not new. Some of the tasks we currently use are quite good in terms of focusing discretely on the particular skill. A good example would be true/false questions for a reading text. Quite clearly none of the other three skills (listening, speaking or writing) are required to complete the task and therefore only reading is being assessed. However, we need to be careful as such a task won’t work for all skills. For example, if you use true/false questions for listening then students either need to read the questions or respond to spoken questions by speaking themselves.

Conclusion

So, what have we learnt?

• When we are talking about assessing a particular skill we should be careful that the task(s) we use do actually assess that particular skill.
• If another skill is required in order to complete the task, this should somehow be reflected in the assessment itself.
• Often a task will require more than one skill, i.e. writing answers to a set of questions for a reading or listening text.
• We need to think carefully about how we weight the different skills used. In other words, if we require written answers for a reading text we must decide on the criteria to assess writing. Furthermore, we need to work out what proportion of the final mark will assess accuracy of the writing (spelling, use of grammar, etc).