The answer is construct validity. It is "the level to which a test or experiment measures what it asserts, or contends, to be measuring." In the classical model of test validity, it is one of three major kinds of validity evidence, along with criterion and content validity.