ABSTRACT
In this chapter the problem of distinguishing between speed and power tests will be considered, and a criterion will be proposed for determining the extent to which a given test approaches a “pure speed” or a “pure power” test. This material is presented as a suggestion toward a differential rationale for speed and power tests. Relatively little has been written on this subject, despite the fact that the problems of item analysis, test length, item difficulty distribution, determination of reliability, and error of measurement are all quite different for the two types of tests. At present most tests are a composite in unknown proportions of speed and power, which makes the development of appropriate theorems in test theory more difficult than for the pure type tests.
