Dynamic changes in reinforcer value: Some misconceptions and why you should care.

Abstract

Abstract The effectiveness of a reinforcer in maintaining behavior (its value) changes systematically with successive deliveries of that reinforcer. Some misconceptions have impeded our understanding of these changes. The misconceptions include the idea that the changes never occur or occur only when large reinforcers are used; that they always occur; that they are always bitonic in form; that they have no theoretical implications; and that they are produced by to the reinforcer. A precise characterization of the factors that alter the effectiveness of a reinforcer is essential for the theoretical understanding of operant behavior and for the use of operant techniques in practice. ********** For many years, operant psychologists have believed that reinforcers lose their effectiveness (i.e., their value) as they are repeatedly presented (e.g., Reese & Hogenson, 1962; Skinner, 1932). This finding has potentially important theoretical and practical implications. Changes in reinforcer value are theoretically important because they are not anticipated by most current theories. As will be argued, the changes may also contribute to the theoretical explanation of a number of operant phenomena, such as behavioral contrast and spontaneous recovery (see Within-session Changes are not Theoretically Important). Dynamic changes in reinforcer value are practically important because operant techniques are often used to strengthen or weaken behaviors. To do this effectively, contingency managers must maintain the effectiveness of their reinforcer or punisher as long as possible. In spite of their potential importance, dynamic changes in reinforcer value were never subjected to an experimental analysis. Instead, they were labeled, satiation (e.g., Reese & Hogenson, 1962). This was regrettable because the only evidence that reinforcers lost their effectiveness with repeated presentation was that operant response rates declined after a large number of reinforcers had been delivered. Factors other than (e.g., fatigue) could have contributed to these declines, but those factors were never ruled out by experimental test. Additionally, even if changes in reinforcer value did produce the declines in response rates, factors other than (e.g., habituation) could have produced those changes in reinforcer value. The study of dynamic changes in reinforcer value was ignored until the rediscovery of the fact that rates of operant responding may not be constant within experimental sessions even when the conditions of reinforcement are constant across the session. Instead, rate of responding often increases, decreases, or increases and then decreases within sessions (e.g., McSweeney, 1992; McSweeney, Hatfield,& Allen, 1990). Subsequent research confirmed that these within-session changes in responding are produced primarily by dynamic changes in the value of the reinforcer For example, probe preference tests show that reinforcer value changes systematically within sessions (McSweeney, Weatherly, & Swindell, 1996a). In addition, several competing explanations for within-session response patterns were ruled out. Rejected explanations include: recovery from handling (McSweeney & Johnson, 1994), anticipation of events that follow the session (e.g., feeding, McSweeney, Weatherly, & Swindell, 1995), changes in a general motivational state (e.g., arousal, McSweeney, Swindell, & Weatherly, 1996a; 1996c), changes in interference from adjunctive behaviors (McSweeney, Swindell, & Weatherly, 1996a) or exploration (Roll & McSweeney, 1997), changes in factors produced by the act of responding, such as muscular warm-up or fatigue (McSweeney, Weatherly, & Roll, 1995; McSweeney, Weatherly, Roll, & Swindell, 1995; Melville, Rybiski, & Kamrani, 1996; Weatherly, McSweeney, & Swindell, 1995), and changes in attention to the task defined in several ways (McSweeney, Roll, & Weatherly, 1994; McSweeney, Weatherly, & Swindell, 1996c; Melville & Weatherly, 1996). …