The need for operational definitions becomes easily apparent when we zero in on variables, like anxiety, that are actually hypothetical constructs, or concepts, which are unseen processes postulated to explain behaviour. Many psychological variables are hypothetical constructs; that is, constructs that we can observe: An ordinarily good student panics and does poorly on an important exam. An actor has stage fright and forgets his lines. From these observations, we infer the existence of “anxiety.” We could say that “anxiety” is a peculiar felling of queasiness that inhibits behaviour. Unfortunately, that definition also neatly fits the first time one of the authors ate row clams. She was not anxious, but she was definitely queasy, and she has not eaten clams since. Furthermore, what we mean by “anxiety” may be quite different if we are talking about a person’s experience before taking a test compared with waiting in line for a roller coaster ride. Clearly, just labelling the variable “anxiety” would be much too vague without a precise operational definition.