In this paper we claim that random behavior may become a legitimate subject matter of a behavioral science. We analyze initially some considerations B. F. Skinner has formulated concerning random behavior, in his Contingencies of Reinforcement. According to Skinner (1969/1975), random behavior "is of no interest as a behavioral process". We present a rigorous definition of "random behavior". We claim that random behavior, so defined, may be functionally analyzed, by examining some empirical investigations concerning random behavior that have been conducted. We also present some empirical and theoretical difficulties that can affect investigations in this area. We conclude that random behavior is a fascinating and important topic to a complete science of behavior.
random behavior; behavioral variability; frequency-dependent selection; percentile - reinforcement schedule