Response variability, a fundamental characteristic of behavior, may be in some cases an induced effect of reinforcement schedules. Research on schedule-induced response variability has shown that continuous reinforcement results in less variability than intermittent reinforcement schedules. Studies on the effects of intermittency of reinforcement, periodicity of reinforcement, and type of schedule have resulted in mixed findings. Contingencies have also been arranged to directly influence operant response variability. These include lag reinforcement schedules, differential reinforcement of novel behavior, differential reinforcement of less frequent behavior, and percentile reinforcement schedules. These procedures are discussed in terms of practicality and implications for use in applied settings. Contingencies and treatment packages that indirectly influence response variability are addressed in terms of response allocation, response generalization, and response covariation. Studies on the effects of a variety of other variables on response variability are also reported, such as levels of food deprivation and drugs. Finally, directions for applied research in response variability are suggested.Traditionally, behavioral variability has been considered an intrinsic, uncontrollable characteristic of behavior or a result of unforeseen, uncontrolled variables (Neuringer, 2002). As a result, the identification of methods to reduce the effects of potential sources of variability has been an important component of much research (Sidman, 1960). Alternatively, behavioral variability itself has been a recent topic of research and can have important implications, both theoretical and practical. From a theoretical perspective, such research has provided behavior analysts with a basis for the investigation of topics considered to be out of the realm of learning theory, such as creativity and originality (Marr, 2003). From a practical standpoint, the identification of conditions that lead to, maintain, or decrease response variability should aid in the refinement of existing practices and the development of new procedures that help reduce stereotypy among the clinical population. Such procedures could 430 LEE ET AL. be used to increase generative language, problem-solving skills, creative play, and normalistic social interactions.Although a behavior analysis of creative behavior is difficult, it is not new. Maltzman (1960) provided one of the earlier reviews of research in which people were taught "original" behavior. More recently, Shahan and Chase (2002) provided a review of behavioral concepts (e.g., adduction) that influence response variability and discussed how they relate to creativity and language. The concepts included stimulus generalization, conceptual behavior, and verbal operants. Neuringer (2002) provided an in-depth review of various procedures, noted their implications for treatment in applied settings, and discussed the sources of variability. Marr (2003) has recently rebutted traditional criticisms of a beha...