Association for Behavior Analysis International

Presenters will discuss the experimental analysis of change.

TBD

When we think about stability of behavior in the context of the experimental analysis of behavior, it has, at least, two meanings. The first is an antonym of instability. Using stable behavior as the criterion for changing experimental conditions is the standard practice of behavior analysts. Thus far, several stability criteria have been proposed and we have seen them in many behavior-analytic references. In our own work, we have experienced sometimes, however, circumstances in which it is difficult to meet a stability criterion. These circumstances raise the question of whether we ever achieve steady or stable behavioral states when experiments extend over a long period. The second meaning of stability is an antonym of variability. We use the terms discrimination and generalization to call a researcher's attention to a specific aspect of the same behavioral phenomenon. We use the terms sensitization and habituation to focus on different effects of the same stimulus. We use the terms stability and variability to emphasize the observer's perspective on the same behavioral stream. For example, some indexes of variability such as the U value have been used in the context of “variability as an operant” studies. Researchers have determined the final stable value of the variability-index by consulting the standard stability criterion, and then compared it with other values produced by other contingencies. But are there any contingency-operations that control stability like variability? Is maintaining an ongoing contingency the only way to achieve this? This point may be related to the issue of a behavioral unit for manipulating by contingencies. Do these two meanings, antonyms of instability and variability, reflect different levels of behavioral status or states? In this presentation, I first will summarize how behavior analysts have defined “stability” and proposed its criteria in their experimental situations. Following this, I will provide some examples from which we can consider the next step in addressing the issues of behavioral stability and its change.

The Heike monogatari reminds us that, “The knell of the bells at the Gion temple/Echoes the impermanence of all things.” So it is also that behavior is fleeting and impermanent. Most of what we study in the experimental analysis of behavior, we do by comparing steady states, ignoring or dismissing what interferes with or occurs between those periods of relatively invariant responding. In this presentation, I will consider what happens as the behavioral versions of the sounds of the Gion temple bells change with the passage of time. Although behavior analysis predominantly has been very usefully concerned with longer-term stable behavior, there is much of interest in the transitions and transient changes that are interspersed within more stable and enduring behavioral phenomena. This presentation reviews some experimental investigations of these more fleeting behavioral phenomena, things like transitions between schedules of reinforcement, extinction, momentary adjustments to dynamic environments, and changes in reinforcement over time.

Behavior dynamics is a field devoted to the fundamental question of what forces drive changes in behavior. Exemplars of different problems and approaches will be briefly reviewed. For example, the phenomenon of shaping is foundational to all of behavior analysis, yet this has received surprisingly little systematic analysis. Part of the difficulty here is the need for careful specification of the initial behaviors which are then subject to transformation into new behaviors through differential reinforcement. Typically, to address this problem, some steady-state "baseline" is established under some simple contingency and then a new contingency is introduced and the transition to a new steady-state performance is investigated. This, too, has received remarkably little attention, so that we understand relatively little about how the typical patterns of responding under commonly studied schedules of reinforcement actually emerge. A more common approach, exemplified by a fair number of studies, is to investigate transitions within sessions by repeated changes in the behavior-consequence relations to see how, for example, response rates track changes in available reinforcement rates. These studies have been subject to the most detailed mathematical modeling. Other approaches include explorations of feedback functions and, in general, complex contingencies wherein ongoing behavior-consequence relations modify how the contingencies control subsequent behavior.

When we consider the interaction of behavior and reinforcers, the temporal relations between the two events are an essential element. To show how reinforcers come to control behavior, much research has been conducted on such topics as response-dependent versus response-independent reinforcement contingencies, immediate versus delayed reinforcement, and so on. These topics are closely related to controversies about the roles of molar versus molecular contingencies and of contiguities versus contingencies. This talk will review various behavior-reinforcer relations over a continuum ranging from the past through the present and into the future. In the long run, each response-reinforcer contiguity constitutes a contingency, and no contingency can exist in the absence of any contiguities. Thus, the question is about the conditions under which contiguities and contingencies work consistently. To address this question, we will compare animal and human behavior as well as positive and negative reinforcement.

TBD

Experts from around the world will present and discuss the current status of behavior analysis in their respective regions.