Association for Behavior Analysis International

The Association for Behavior Analysis International® (ABAI) is a nonprofit membership organization with the mission to contribute to the well-being of society by developing, enhancing, and supporting the growth and vitality of the science of behavior analysis through research, education, and practice.


32nd Annual Convention; Atlanta, GA; 2006

Event Details

Previous Page


Symposium #292
International Symposium - Recent Advances in Understanding “Reinforcement”
Monday, May 29, 2006
9:00 AM–10:20 AM
Hong Kong
Area: EAB; Domain: Basic Research
Chair: Randolph C. Grace (University of Canterbury)
Discussant: Randolph C. Grace (University of Canterbury)
Abstract: Over the last 6 years, the study of preference in concurrent schedules has moved from extended analyses (the generalized matching relation) to local analyses. This move has been driven largely by the collection of massive amounts of data. Based on these data, Davison and Baum (2005) attempted an initial synthesis, and questioned the utility of the notion of “reinforcement”. In this symposium, we add further, targeted, data to the mix, we add further analyses and interpretations of post-reinforcer preference pulses, and we propose a new, and more general, way to think about the way in which the organism and environment interact to produce mutual change and adaptation.
Preference Pulses: New Data, New Questions.
MICHAEL C. DAVISON (University of Auckland)
Abstract: What effect do additional, response-contingent, but uncorrelated stimuli have on preference in concurrent schedules? What happens to preference pulses if delays in blackout are arranged after reinforcers on concurrent schedules? What happens to preference pulses if only one of the two concurrent schedules runs after reinforcers? Results from three experiments will be briefly described and their implications for notions of reinforcement—and for the correlation machine—are discussed.
Preference Pulses: New Analyses, New Questions.
DOUGLAS ELLIFFE (University of Auckland)
Abstract: Since 2000, about a dozen papers by the Davison/Baum/Elliffe/Landon/Krägeloh research group have reported preference pulses, or strong but transient preference for the just-reinforced alternative on concurrent VI VI schedules. Pulses in this literature have plotted log response ratio either as a function of time since reinforcement or of number of responses since reinforcement. This paper examines which analysis better captures regularities across experiments and consistent effects of independent variables, and explores a possible analysis that includes both response position and time information. The literature to date has discussed pulse effects in a relatively descriptive, verbal, and qualitative manner. While this has led to theoretical advances, as exemplified by this symposium, only large and unambiguous effects have been able to drive those advances. This paper also considers whether the time is ripe for a rigorously quantitative description of pulses, so that more subtle pulse effects can be unequivocally shown. If we can approach the quantitative precision of, for example, generalized-matching analyses of choice, rapid theoretical advance will become more likely.
The Correlation Machine: A New Way to Understand Reinforcement, Punishment, and Stimulus Control.
WILLIAM M. BAUM (University of California, Davis), Michael C. Davison (University of Auckland)
Abstract: During its 100-year history, the law of effect has been challenged in two ways: (a), by questions about the necessity of reinforcement for learning; and (b), by questions about the sufficiency of contiguity as the key factor for learning. Examples of the first type of question include sensory preconditioning and latent learning. Examples of the second type of question include Rescorla’s discussion and demonstration of the need to replace contiguity with contingency and Baum’s proposal of the correlation-based law of effect. The idea of the organism as a correlation machine replaces the contiguity-based law of effect and accommodates all of the earlier criticisms. It offers a powerful framework within which to understand the older data and recently gathered data that argue against the traditional ideas of reinforcement, punishment, and stimulus control.



Back to Top
Modifed by Eddie Soh