The Missing Cornerstone: A Model of Cognition
The model of cognition “refers to a theory or set of beliefs about how students represent
knowledge and develop competence in a subject domain” (National Research Council
2001, p. 44). Describing what he calls standard test theory, Mislevy (1996) argues that most
educational measurement lacks such a model of cognition. Instead, standard uses of
classical test theory, generalizability theory, item response theory, and factor analysis or
structural equation modeling are bound up in a paradigm of behavioral psychology in
which the target of interest is “a summary of a behavioral tendency in a class of stimulus
situations—an overall proficiency in the prescribed domain of tasks” (Mislevy 1996, pp.
386–387). As a consequence, the valid interpretations of test scores are limited:
Through the use of standard test theory, evidence can be characterized and brought to
bear on inferences about students’ overall proficiency in behavioral domains, for
determining students’ levels of proficiency, comparing them to other students or to a
standard, or gauging changes from one point in time to another. Conjectures about the
nature of this proficiency or how it develops falls largely outside the mental
measurement paradigm’s universe of discourse (Mislevy 1996, p. 388)
The Missing Cornerstone: A Model of CognitionThe model of cognition “refers to a theory or set of beliefs about how students representknowledge and develop competence in a subject domain” (National Research Council2001, p. 44). Describing what he calls standard test theory, Mislevy (1996) argues that mosteducational measurement lacks such a model of cognition. Instead, standard uses ofclassical test theory, generalizability theory, item response theory, and factor analysis orstructural equation modeling are bound up in a paradigm of behavioral psychology inwhich the target of interest is “a summary of a behavioral tendency in a class of stimulussituations—an overall proficiency in the prescribed domain of tasks” (Mislevy 1996, pp.386–387). As a consequence, the valid interpretations of test scores are limited:Through the use of standard test theory, evidence can be characterized and brought tobear on inferences about students’ overall proficiency in behavioral domains, fordetermining students’ levels of proficiency, comparing them to other students or to astandard, or gauging changes from one point in time to another. Conjectures about thenature of this proficiency or how it develops falls largely outside the mentalmeasurement paradigm’s universe of discourse (Mislevy 1996, p. 388)
การแปล กรุณารอสักครู่..
