An important approach to understanding child behavior and development is the experimental analysis of such behavior. The experimental analysis procedure must be distinguished from related analyses used occasionally. An analysis by anecdote is an analysis based upon the accumulation of recurring associations; for example, B followed by A. This does not assure the existence of a causative relationship. An analysis by correlation is a survey of two anecdotes, (1) if B, then A, and (2) if no B, then no A. But this analysis does not assure that when factors 1 and 2 exist, some factor 3 exists or that both A and B are controlled by C. The experimental method requires that the experimenter manipulate or control A and B in an arbitrary fashion. This reasonably precludes control by some unknown C and reasonably illustrates the causative relationship. The age of a child limits the application of the experimental analysis approach; that is, very young children are generally not available for use in a comprehensively controlled environment. The operant behavior procedure does not require control over a broad range of environmental factors, however, and evidence from such procedures indicates the possibility of investigating child behavior and development through the use of reinforcement, punishment, and extinction contingencies. (WD)