Language testing has improved parallel to advances in technology. Two basic questions in testing; Can language be broken into skills and components? Are all skills attained separately at different rates or they are attained together at the same rate?
There are four principal coexisting but competing trends in language testing. Prior to the study of language as a science language testing was characterized by the lack of a well-established theory=grammar translation method.
A- Attempts to assess language abilities through different subtests that sample distinctive elements from each skill, aspects of the skill, and a component of the skill. Structural linguistices+behavoral psychology =discrete point approach.
1-Short multiple tests which could be easily verified for reliability. 2-Each item samples a particular element through the use of one skill. 3-Items are not dependent one another, chancing one item does not change the testee’s performance on the other items of the test. = TOEFL
1-DP tests are the ideal form of tests for diagnostic purposes. They furnish results which are radially quantifiable
1-They are difficult to construct. 2-It is difficult to determine which point to test and which not to. 3- They cannot portray the subject’s ability to use language in real life situations. 4-Contrastive analysis has been criticized as being short-sighted.
DP approach will continue to be the major tradition in test construction for years to come.
Involves the testing of language in context and thus primarily concerned with meaning and the total communicative effect of discourse. Cloze, dictation, composition writing, and oral interview. Some believe that DP and IA form the two extremes of one continuum.
1-Time consuming (composition writing) 2- The reliability and validity of dictation and close tests is in question. They do not meet the criterion of predicting or assessing a unified and integrated underling linguistic competence. Language tests should assess the efficiency of the expectancy grammar of the learner.
The psychologically real grammar a learner internalizes when he learns a language. The measuring instrument developed to tap this efficiency has been called pragmatics. Question answering dictation, narration, noise test, translation, composition writing, oral interview and paraphrase recognition are examples of pragmatic testing.
Expectancy grammar resembles the type of grammar required by language users in the execution of a speech act. It has been claimed that pragmatic tests attempt to assess the ability to recognize relationships holding between linguistics elements and extra linguistic context in which elements are appropriately used.
A pragmatic test should fulfill two criteria: 1-It should invoke and challenge the subject’s developing grammatical system. 2-The test should involve linguistic sequences that relate to extra linguistic context in meaningful ways. Not all the integrative tests fulfill both of the above criteria.
1-Communicative competence is so global that typical pragmatic tests cannot adequately measure it. 2-The close procedure is regarded as a test of language usage. 3-Dictation is also criticized because it does not give any convincing proof of the examinee’s skill to actually use the language.
Both communicative approach and integrative approach emphasize the importance of the meaning of utterances, rather than their form and structure.However,communicative tests are concerned primarily with how language is used in communication.
Perhaps the most important criterion for communicative test is that they should be based on precise and detailed specification of the needs of learners for whom they are constructed.
The attempt to measure different language skills in communicative test is based on a view of language referred to as divisibility hypothesis.
Finally,communicative testing has introduced the concept of qualitative modes of assessment in preference to quantitative ones. What are the benefits of qualitative evaluation? Each students’ performance is evaluated according to his or her degree of success in performing the language task and it is not compared with other students
1.Norm-referenced It gives meaning to raw scores by permitting the comparison of test performance of examinees with those of other representative subjects on a local, regional or national basis These tests are designed to yield maximum discrimination among testees and spread their scores along a normal probability curve.
When a great number of testees miss an item it is eliminated from the test.This approach is useful in aptitude testing where the purpose is to make differential predictions Norm referenced test yields a rank order of candidates based on distribution of scores.
2.Criterion-referenced interpretation: It refers to what testees can do with what he knows, not on how he compares with others.it does not matter in principle whether all candidates are successful, or non of the candidates is successful.
Characteristically, criterion-referenced tests are designed before the instruction is designed. A criterion is set in advance and those who do not meet the criterion are required to repeat the course. Selection tests and proficiency tests fall into the realm of criterion-referenced testing.