Foreign Language Teaching Theories外语教学理论

Foreign Language Teaching TheoriesChapter 7 Language Testing

7.1 Purposes of language testing

As in any discipline, evaluation/testing plays an important part in the teaching and learning of a second or foreign language. Testing is an essential component of any ESL/EFL instructional programme. It provides the means of assessing students’ overall proficiency in the language, of identifying specific problems students are having in learning the language, of placing students at an appropriate instructional level, and of indicating areas in which teaching may not have been effective. Testing can be an effective means of teaching, and it often provides a strong motivation for learning.

Summative evaluation and formative evaluation Testing does not only have the purpose of measuring the language behavior of inpiduals, but it is also useful instruments for evaluating programmes by means of an analysis of the results of an entire group of learners. Summative evaluation: the evaluation that takes place at the end of a programme, for evaluating a programme as a whole; a final test used for measuring what an inpidual has learnt in the whole programme. Formative evaluation: the evaluation that takes place during the programme, intended to adjust or guide the programme; to determine what the learner has or has not learnt, so that the programme can be adapted to the specific needs of the inpidual learner.

Placement tests: tests used to determine the initial level of proficiency of learners who have enrolled for a particular course in order to be able to place them at an appropriate level in a programme. Aptitude tests: tests intended to provide information on the likelihood of success or failure of a learner in language learning These tests consist of different tests which measure such abilities as (a) sound-coding ability (b) grammatical coding ability (c) inductive ability (d) memorization (rote-learning) ability

Achievement tests: tests used to measure how much of a language a learner has learnt with reference to a particular course of instruction (progress tests, final tests) Diagnostic tests: tests designed to specify the learner’s particular strengths and weakness, to show what skills or knowledge a learner has or has not mastered, and ideally speaking, to help discover the causes of failure.

Achievement tests answer the question how much the learner knows, while diagnostic tests answer the question what he/she knows and what he/she does not know, and should be able to answer the question why he/she knows or does not know something. General proficiency tests: tests used to indicate what an inpidual is capable of doing now with an L2, whether an inpidual is sufficiently proficient in an L2. Such tests are usually standardized for world-wide use.

Harris (1969:13): All good tests possess three qualities: validity, reliability, and practicabil

ity. that is to say, any test that we use must be appropriate in terms of our objectives, dependable in the evidence it provides, and applicable to our particular situation. Harris’ third quality is especially important from the teacher’s point of view. Facility in administration (no need for special pencils or equipment, for example) and ease of scoring and factors to consider in a testing programme.

7.2 Some general requirements for tests

7.2.1 Reliability: the degree of consistency ( internal consistency and consistency over time ) shown by the results of a test. A test is said to be reliable if it gives the same results when it is given on different occasions, or when it is marked by different people.

Reliability has much to do with the objectivity of tests: First, test conditions and tasks have to the same for all participants (i.e. conditions and tasks have to be standardized). Second, the instructions for the judges have to be such that different judges, when marking a independently of each other, come up with the same evaluation. Third, the test results should not leave room for different interpretations.

Stability reliability and test-retest methodA measuring device is stable if it gives the same results when administered twice on the same testees, assuming there is no change in the testees between the two measurements.

Given that the testers and the testees, and the test situations are not varied, we can measure the reliability of a test by testing and retesting the same testees at two different times, and then correlating their scores, -- the higher the correlation between the two tests the higher the reliability. This is called test-retest method which indicates stability reliability or test-retest reliability.

Equivalence reliability and the parallelforms method A measuring device is equivalent to another they produce the same results when used on the same testees. This is to administer parallel or equivalent forms of a test ( with the same kind of items of the same degree of difficulty ), called the parallel-forms method, which indicates the equivalence of the two forms of the test, called the equivalence reliability.

If testees get similar scores on parallel forms of a test, this suggests the reliability of the tests. If the correlation between the two tests is high (i.e. the results correspond closely to each other), then the tests can be termed reliable.

In estimating internal consistency, a simple procedure is the split-half method. The test is pided into two halves, usually into odd and even number items, and the scores of these two sets are correlated.

But there is a problem: if items are graded according to increasing difficulty, pision according to odd and e …… 此处隐藏：4152字，全部文档内容请下载后查看。喜欢就下载吧 ……