Testing the spoken competence of learners in gimnazjum

Numer JOwS: 
str. 67

Speaking is one of the most challenging skills, both for the foreign language learner to carry out and for the teacher to test. When speaking, learners have to make many split-second decisions about what to say and how to say it, choosing from a store of language which may still be very limited.  Add to this the anxiety they feel when they think someone is judging them in a testing situation and the simple problem they have of thinking what to say and we begin to realize the scale of the problem. Designing a test and deciding how it should be rated consequently needs to take into consideration not only linguistic but also psychological factors.

Pobierz artykuł w pliku PDF

This article gives a brief overview of a test of speaking which was conducted in March 2011 among 499 learners from the final class of gimnazjum.  The test was conducted with some of the learners who took part in the tests of English for the first European Survey on Language Competences.  The speaking test was not part of the official test, but was designed and carried out only in Poland, with a view to getting a picture of the level of skills learners have in English at the end of Stage III of the school system.

Pobierz artykuł w wersji polskiej

Spoken competence

Speaking, an element of communicative competence, can be divided into two main skills: spoken interaction, where two or more speakers engage together; and spoken production, where one speaker attempts to convey an idea, give information, tell a story etc. Performance of each of these two main skills involves other skills such as selecting the appropriate words to use, knowing how to pronounce them correctly, selecting grammatical structures and functional language. In the case of interaction it is necessary to listen to what the other person (or people) say, understand their message and intentions and to react appropriately, selecting not only the language needed, but also taking into consideration other social factors such as level of politeness, and cultural appropriateness.  In spoken production the speaker has to craft their speech, organizing the order of what to say; connecting sentences together logically.  Additionally, all of this must be done under immense pressure of time, as spontaneous speech takes place in ‘real time’. 

For the foreign language learner attempting to speak in real time is challenging. Research suggests (Skehan 1998) that three areas compete in speech production: fluency, complexity and accuracy, and that focus on one aspect impacts on performance in the other areas. For example, if a learner still finds it difficult to find the words they want to use quickly and without effort and yet decides that they must express these words in order to convey their ideas (focus on complexity), then the fluency of their speech will suffer, as they need time to find the words they want. Similarly, the learner who sets themselves the target of speaking without mistakes and who still needs time to think which structures to use when, or how to form the necessary part of a verb, will also show a lack of fluency in their speech.  By contrast the learner who aims to get their message across without concern for accuracy or exact choice of words will speak with fluency but may demonstrate lower accuracy and complexity in their speech.

Affective factors

Speaking in a foreign language, as we have seen, is complex and also potentially frustrating.  The learner who has not reached the level where they are able to say what they want with reasonable ease may feel that what they are able to produce in the foreign language does not reflect themselves as a person, or the ideas they want to convey.  They may complain of sounding childish, that their interlocutor will think them uneducated, or of limited intelligence, or they may simply give up and decline to speak or shorten their response to a minimum.  In this situation the attitude and approach of the person they are speaking to is key.  If the interlocutor is patient, listens empathetically, nods and encourages and shows interest in what is being said, rather than how it is being said, then the learner will most likely do their best to continue.  If, however, the listener shows impatience, fidgets, drums their fingers on the table, or sighs heavily, embarrassment will cause the learner to hesitate more, slow down, repeat themselves and eventually breakdown and stop speaking altogether. There will most likely be a similar effect if the listener starts to interrupt and correct the learner’s attempts to speak.  The supportive interlocutor may gently and diplomatically respond to ‘appeals’ from the learner, by quietly supplying them with words they appear to be searching for, or by completing their sentences if there is a breakdown and in this way enabling the learner to convey their message. Past experience may influence how the learner feels about speaking.  If they feel that past attempts have been humiliating then this may block the desire to communicate in the foreign language.

Testing spoken competence

When designing a test of spoken competence all of the factors described above have to be taken into consideration.  We must expect that learners will be anxious and that this will, at least at the beginning of the test, impact on their performance. Research suggests (Foster and Skehan, 1996) that allowing a short time for preparation before being asked to speak will help learners feel more confident.  Similarly clear instructions are a must. If we want to get a overall picture of the learner’s level of spoken competence then we need to include a variety of tasks, testing in turn interaction and production. Care needs to be taken that the tasks we choose are not  simply designed to elicit an element of language, such as past tense, but also that any interactions take place in situations in which our learners might find themselves and where they would genuinely need to speak English.

For testing interactive competence the learner needs to speak and respond to someone, either another learner or the teacher, or a simulated ‘person’ online. If we decide to test two learners together we need to consider how to organize pairs so that learners can cooperate together and not compete.  We want learners to help their partners achieve a task and not to make it more difficult for them! There is the additional problem that learners may subconsciously mirror each other’s speaking, borrowing words and phrases from each other and even their partner’s mistakes. One  person in the pair may dominate and so sometimes it is hard for the other person to demonstrate what they are really capable of. Clear instructions and an opportunity to watch demonstration tests can help to overcome these difficulties.

When testing spoken production we need to consider how we intend to stimulate the learner to speak and what kind of prompt to use.  This may be a picture to describe, a topic to discuss, a story to tell from pictures, or alternatively an open question given by the teacher or examiner which the learner is expected to answer. The topics should be suitable for the age group, taken from subject areas listed in the core curriculum and be presented in such a way as to encourage learners to speak. With teenagers, and most particularly with boys, it is sometimes challenging  to get them talking, even in their own language. Here how the interlocutor behaves will be key.  Long wait time is needed after the question is asked, together with encouraging non-verbal signals. In general, first we need to build the learner’s confidence by asking easy questions on familiar topics, before starting to challenge them a little more. As the interlocutor role is key to ensuring we elicit a representative sample of the learner’s performance skills in English, the ideal situation is where there are two people present during the test: one to speak to the learner and the other to rate the performance. However, as this is costly and logistically difficult an alternative is to make a recording of the test to allow rating to take place after the test under less stress.