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.
