“a test is valid for anything with which it correlates” (p. 429). The idea here was that, if there were a “gold standard” of the con-struct (often, an earlier test), and if scores on the test correlated with that gold standard, the test could be inferred to be a measure of the construct. Note that this is a variety of the “if it walks like a duck and quacks like a duck, it must be a duck” argument.