ABSTRACT
This chapter discusses the classical test model, which splits the test score into the sum of the true score and a random measurement error. The two basic concepts in classical test theory (CTT) are reliability and validity. I discuss the definition of test-score reliability, and I notice that test-score reliability is a group characteristic that should be distinguished from the measurement precision of individual test scores. I discuss several well-known methods for estimating test-score reliability. Next, I discuss the concept of measurement precision, and I explain how to statistically test whether a given test score differs from a cut score, from another person’s test score, and whether a person’s posttest score differs from his/her pretest score so as to establish whether a treatment was successful after a specified, theoretically justified period of time. Finally, I discuss that validity refers to the attribute the test measures but also to the purposes for which a test score can be used. I argue that both aspects of validity are important and that one cannot be ignored without impairing the other.
