face validity | content validity |
Face validity is a very informal judgement. It simply means that the people sitting the test, the people giving the test, and others affected by it such as parents, employers, and government officials see the test as fair and reliable. A reliable test which may have good content and predictive validity may be so different from what the public expect or consider relevant that its poor face validity is enough to stop it being used. Good face validity is not a guarantee of reliability or other kinds of validity. | Content validity involves considering whether the content of the test reflects the content of the skill, language, or course being tested. For example, in order to decide if a test of academic listening skill has content validity, we would need to decide what are the components of the academic listening skill and how is this skill used. We might decide that academic listening involves note-taking, dealing with academic vocabulary, and seeing the organization of the formal spoken discourse. |
Quantitative approaches | Qualitative approaches |
Quantitative approaches tend to analyze data in terms of pre-existing categories and the researcher then seeks to investigate the nature of these items in the data. For something as dynamic and socially grounded as spoken discourse, this use of pre determined categories can be unhelpful. The strength of the qualitative paradigm is that it works from the ‘inside’ of instances of talk towards patterns and regularities and is able to uncover aspects that the investigator may not have imagined existed. | A widely used method among the qualitative approaches to researching speaking is conversation analysis (CA). This method puts high value on the careful analysis of examples of real (i.e. non elicited) talk to understand how speakers create meaning and organize their discourse as social action. The CA analyst is interested also in what linguistic resources (syntax, prosody, gaze, laughter, silence, and so on) speakers use to ‘do talk’ and how these are different in specific varieties of language and discourse contexts. The ‘pure’ CA approach, therefore, is unique in that it seeks to understand the nature of speech primarily from observation of nonelicited data and through this process gain insights about broader patterns and meaningful regularities appearing in the interaction. |
Copyright (c) 2021 vulmsexpert All Right Reseved
0 Comments