Validity for What? The Peril of Overclarifying

摘要

As Newton (this issue) so ably demonstrates, the concept of validity is both important and problematic. Over the last several decades, a consensus definition of validity has emerged; the current edition of Standards for Educational and Psychological Testing notes, "Validity refers to the degree to which evidence and theory support the interpretations of test scores entailed by proposed uses of tests" (AERA/APA/NCME, 1999, p. 9). The two keys to this decision are that the term validity refers to the interpretations of test scores, not to the tests or the test scores themselves, and that the inferences that are validated are a product of the proposed uses of tests. Newton shows that this definition is potentially confusing, in part because of the multiplicity of uses of tests and the consequences of using tests in various ways. He proposes a definition of validity that is both more narrow and more precise, focusing on whether there is sufficient evidence for interpreting the results of an assessment-based-decision-making procedure as reflecting the attribute entailed by the decision.