Testing English grammar

Selection items require students to choose the correct answer from several provided options. The manner of scoring is usually dichotomous (0 points or 1 point), however, sometimes partial credit scoring (0, 1, or 2 points) may be applied (Brown, 2003). The most commonly applied selected response tasks include multiple-choice, discrimination, noticing, and matching/pairing tasks, which will be discussed in the subsequent paragraphs

1. Multiple-choice tasks
In multiple-choice tasks, a test-taker is supposed to select the correct response from presented options. Typically, there are four of them, but variations on the number of options are possible. One of them is the correct answer, while the other ones are incorrect, being the so-called distractors, whose goal is to move test-taker’s attention away from the correct response (Coombe et al., 2007). This is what usually a multiple-choice task, aimed to check the knowledge.

This technique is one of the most widely applied as far as testing grammar is concerned, particularly in standardized testing, most probably because it is easy to administer, and even easier when it comes to scoring, which can often be done by computers. Moreover, they are universal, and can be applied in testing on all levels of education, and students are familiar with this form of testing. If the items are well-designed, their reliability is high as well. Finally, students’ performance is not affected by their writing ability, as nothing more than ticking the correct answer is expected.
However, one of the greatest disadvantages of multiple-choice format is the fact that it promotes guessing, and test-wise students can take advantage of it, which can affect test results. Cheating is made easier as well. Furthermore, they may not be considered as an authentic language use, and are not really applicable in testing productive skills, such as speaking. In addition, some teachers tend to use this format almost exclusively, which may exert a negative influence on classwork (Brown, 2003; Coombe, et al 2007; Madsen, 1983).
Purpura (2004) also mentions multiple-choice error identification task, in which students are presented with sentences, or paragraphs, containing one grammatical error, which they are supposed to identify.

2.Discrimination tasks

In this type of testing tasks, learners are expected to pay attention to the input, and respond to it by choosing between contrasts or opposites, such as true/false, yes/no, or right/wrong, with the first example being the most frequent. Questions are typically formed as statements, which students have to assess as true or false.
Discrimination tasks are attractive for test designers for several reasons. By means of such items, one can test large amounts of content in a relatively short time, as they need less time to be answered. In addition, scoring is quick and easy. However, like in the case of multiple-choice questions, true/false questions promote guessing, and the chance of getting the correct answer is 50%, which may affect the results of the test. To diminish this guessing factor, test developers may sometimes include another option, such as “not given.” Additionally, some of them require students to correct false statements, or find passages in the text which justify the given answer (Coombe et al., 2007).

3.Noticing tasks / consciousness-raising tasks
This type of tasks includes a wide range of linguistic/non-linguistic input. Test-takers must pay attention to form or meaning, or both, and thus consciously realize the existence of certain language features in English. Students are required to underline, or circle, a specific language feature in the text sample.

4.Matching/pairing tasks
This type of task typically includes a short conversation, that is a stimulus – a question or a statement – and a response. A test-taker is supposed to select correct answers to the stimuli. Matching/pairing tasks are helpful in measuring student’s knowledge of grammatical and pragmatic meaning, although the knowledge of grammar forms may be useful in successful completion of these tasks.

5.Limited production tasks
Inlimited production tasks, test-takers are asked to provide a response which includes a limited amount of language production, be it a word or a sentence. In the case of this type of test items, scoring can be dichotomous, if it is assumed that students can provide fully correct or fully incorrect answers, or partial credit, which allows also for partial correctness. The most popular formats as far as testing grammar is concerned are gap-filling, short answers, or dialogue-completion tasks.

6.Gap-filling tasks
This kind of testing involves a sentence, a dialogue, or a text passage, in which some words are deleted. Basing on the context, students are required to provide response which fills the gaps. In terms of grammar, gap-filling tasks may be focused only on form, or both form and meaning. The latter case is obviously more difficult, as not only the knowledge of appropriate grammatical form is required, but also a correct lexical item should be inserted.

7.Short-answer task
These tasks include a reading passage or other stimulus, followed by a question, or series of questions, to which a short response, typically in the form of a word or a sentence, should be provided.

8.Dialogue-completion tasks
A short dialogue is provided, in which some phrases are deleted. A test-taker’s task is to insert the missing parts of the conversation, which is accurate in grammatical terms. As in the case of other short-answer tasks, it can be designed to check the knowledge of grammatical form, or a form and meaning.