Measurement when the data fit; diagnosis when it doesn’t
Our purpose when undertaking this venture was not to explain data or even to build better instruments. It may not seem like it based on the discussion so far but our objective is say something useful about the objects. Although the person and item parameters have equal status in our symmetrical model, they aren’t equally important in our minds. The items are the agents of measurement; the person is the object.
Continue . . . Vc. Diagnosis with and from the Model