ABSTRACT
This chapter discusses the principles of reliability and its relationship to validity. It illustrates an important principle: A test with high reliability may have low validity. The chapter presents an example of this principle: An employer wants to reward the best employees with end-of-year bonuses. The employer decides that to be perfectly fair, she should use a completely objective method for determining who should get the bonuses. The employer examines the employees' time cards and selects those who were never late for work during the previous year to receive bonuses. When evaluating measures, validity is more important than reliability. This should be clear from considering the employer basing bonuses on employees' time cards. A complex measure involving subjective judgments of employees' performances that taps a variety of important types of behavior and achievement on the job would be much more valid than a highly reliable measure that considers only punctuality measured by examining time cards.
