A test contracting firm has agreed to create a three part social studies test intended to measure sixth graders knowledge of geography, government, and history. The test is to be used as a statewide assessment to govern the allocation of staff development funds for elementary teachers. Because of the limitations in administrative time, only 15 items are allowed for each of the three content areas. If the test development firm wants to know if its three test subsections are measuring, with suitable consistency, what they are supposed to be measuring, what kind of reliability evidence would be best?