This presentation observes the possibilities that language tests grow language abilities or that language abilities are measured through language tests correctly.  Firstly, there is a question that language abilities and pedagogies in classrooms might be advanced at the exercises of language tests as the item response theory (Hambleton, 1991).  Tests are relied on the items and the test scores are said to be one of the language abilities of test takers (Bachman and Kunnan, 2005).  It would be a negative side in second language acquisition, which is dependent on the grammar instructions (Ortega, 2009).  Secondly, it is a main concern for researching in language tests to define the way to measure language abilities.  These might be based in the distinction of language competence and language performance (Chomsky, 1965), which has developed the structuralism.  The latter might have the assessment in the level of performance.  With the ideal measurement, logically language abilities could be improved.  To this end, those observations suggest that there is a doubt in relations between language tests and language abilities.  However, now it is not examined through scientific researches, but through the investigation of tests.  If the language tests are correct, the language learners with high scores in grammar exercises should manipulate the language.  This presentation indicates that language tests and language abilities are not based on pedagogies in classrooms.  This research includes the warning to the pedagogies on language tests, which do not become to investigate SLA researches without tests.

Testing and evaluating,Pedagogy,Performance-based design