Testing And Validity Essay

Based on Messick (1996), validity is defined as "an overall evaluative judgment of the degree to which empirical evidence and theoretical rationales support the adequacy and appropriateness of interpretations and actions based on the test scores or other methods of measurement" (p. 221). Therefore, validity is the interpretations of test scores. In fact, validation is not directly related to test, however it's related to the inferences and interpretations regarding the test scores. It is said that washback is inextricably bound up with test scores' interpretations and inferences. According to Messick (1996) and Hamp-Lyons (1998) washback is considered as subcategory of construct or consequential validity. The use of test results, the impact of test use on teachers and learners, inferences from test scores, tests use and abuses, and intended and unintended consequences of score inferences (Hamp-Lyons, 1998), are all assumed as parts of construct validity (Messick 19881; 1989 as cited in Shohamy, 1996). Thus, all tests have washback effect, and test validity refers to the fact that whether the test leads to positive or negative washback. To minimize or discard the negative washback, we must reduce construct under-representation and construct irrelevancy. For instance, the

Consequential validity includes, tests effects on test-takers and teachers, score interpretations and misuses, and unintended tests' use. Messick (1996) considers both convergent evidence and discriminate evidence in test validity and assumes to relate the inferences to convergent evidence to support them and to discriminate evidence to discount them. Chappelle (1999) accepts Messick's theory and assumes validity as integrated quality being coherent due to the evidence of test scores'