ABSTRACT
Reliability has been regarded as the correlation of a given test with a parallel form. Correspondingly, the validity of a test is the correlation of the test with some criterion. In this sense a test has a great many different “validities.” For example, the ACE Psychological Examination has one validity for predicting grades in English and a different validity for predicting grades in Latin. It is also found in studying various validity coefficients for a given test that they vary from school to school, and from time to time. In other words, validity cannot be regarded as a fixed or a unitary characteristic of a test. As new uses for a test are contemplated, new validity coefficients must be determined; and, when use of a test is continued, the validity coefficients must be redetermined at intervals. In the remainder of this chapter we shall refer to “test validity” only in the sense that we are considering the relationship between test length and its validity for predicting a specified criterion. In most practical investigations of a test, we should be comparing several different validity coefficients.
