• Version
• Download 342
• File Size 229.77 KB
• File Count 1
• Create Date August 2, 2018
• Last Updated August 2, 2018

Critical Studies of the Notion of "Language Ability" in Bachman and Palmer’s Test Model

This paper presents a critical discussion about the relationship between “,language ability”, in Bachman and Palmer‟,s AUA (Assessment Use Argument, 2010) test model and the common concept of language ability. The purpose is to point out that although the definition of “,language ability”, in the influential AUA model is thought-provoking, its application would seem to be almost impossible in practice, especially regarding standardized tests. It is important to realize that the common concept of language ability (four language skills) is not the main concern in the AUA model. This paper claims that the test validation should be more focused on the observable test performance, i.e. language proficiency, rather than the psycholinguistic “,competence or ability”,. As limited by space, the empirical analysis of the test scores (N=78, using Paired-Sample T Test and Pearson test) is just presented briefly to consolidate the author‟,s claim. Generally, it is important for test designers to realize the limitations of the AUA model in practice. Rather than the unobservable competence, it is test performance to which we need our interpretations about language ability, i.e. language skills, to generalize. ,Keywords: AUA test model, language ability, test performance, competence

Attached Files