3. Measurement 131
3 Measurement The reason that both naive native speakers and language testing raters tend towards a holistic strategy is that evaluation is inevitably a comparative pro- cess, and hence: "Evaluations fall along the same continuum, allowing for meaningful com- parisons across persons, tasks, persons and the like. The universal nature of this general evaluation has been noted in several separate research lit- eratures .... The existence of a single and universal dimension implies that the consistent scaling of performance is at least theoretically possible." (Murphy and Cleveland 1992: 11 9-120) \\!hen Scriven, inventor of the term formative evaluation (Scriven 1967; cited in \\!iddowson 1990: 51), was asked if evaluation was always compara- tive, he is reported to have replied "No, only good evaluation is compara- tive" (cited in Glass 1978: 259). This is arguing against the simplistic divi- sion into master / non-master; competent / incompetent which has domi- nated thinking in criterion-referenced assessment: "knowing and being able to are not absolute, ali-or-nothing attributes" (Trim 1978: 51). Criterion-referenced Assessment An intuitive definition of criterion-referencing is that students' perform- ances are judged in relation to a defined standard and not in relation to their peers. The standard defines where the learner is on the continuum of learning. Thus as he/ she learns he/ she progresses through standards and makes visible progress even if his/her position in the class (e.g. 7th) remains unchanged. This is in opposition to norm-referencing, which defines a stu- dent's place in relation to his/her peers: rank in class...
You are not authenticated to view the full text of this chapter or article.
This site requires a subscription or purchase to access the full text of books or journals.
Do you have any questions? Contact us.Or login to access all content.