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.


43rd Annual Convention; Denver, CO; 2017

Event Details

Previous Page


Symposium #35
Stimulus Control: Testing the Boundaries
Saturday, May 27, 2017
10:00 AM–11:50 AM
Hyatt Regency, Centennial Ballroom B/C
Area: EAB; Domain: Basic Research
Chair: Vikki J. Bland (The University of Auckland)
Discussant: Jennifer Sutton (Brescia University College)
Abstract: When behavior is reinforced in the presence of a stimulus, that stimulus may come to control the rate or probability of that behavior (stimulus control). In experimental and applied settings, the extent to which stimulus control is established or maintained depends on a number of factors, such as the similarity between training and applied environments, the presence of other stimuli, the strength of the learned association between the stimulus and consequences, and which element of the stimulus controls behavior. This symposium will present new research examining how manipulating these factors impacts stimulus control in pigeons, humans, and rats. Presentations will discuss whether a negative discriminative stimulus can function as a punishing consequence; explanations for why some elements of a multi-element stimulus may selectively control behavior; whether stimuli that provide conflicting information can jointly control behavior; and how differences in stimulus control between rat species may be applicable to rodent models of Attention-Deficit/Hyperactivity Disorder. The findings presented in this symposium are likely to be relevant to both experimental and applied researchers interested in extending the boundaries of stimulus control.
Instruction Level: Intermediate
Keyword(s): compound stimuli, punishment, stimulus control, timing
Punishment Revisited: Stimulus Control and Behavior Reduction
VIKKI J. BLAND (The University of Auckland), Sarah Cowie (The University of Auckland), Douglas Elliffe (The University of Auckland), Christopher A. Podlesnik (Florida Institute of Technology)
Abstract: Punishment describes a reduction in the future probability of behaviour, due to a stimulus or event produced by that behavior. Although punishment is a real world process requiring ongoing research and improved understanding, the use of punishing stimuli is generally avoided due to ethical concerns. However, excluding punishment from behavioral treatments is problematic when alternative methods fail to reduce harmful behaviour to safe levels. The present study investigates whether a negative stimulus (S-) will function as a punishing consequence for behavior maintained by positive reinforcement. Five pigeons are used. Following training, one stimulus (S+) predicts response-contingent food deliveries on a variable interval schedule. Simultaneously, an S- stimulus previously associated with the absence of food is presented on a variable rate schedule. Food deliveries are not withheld when the S- stimulus is presented. Results show that the overall rate of key pecking for food is reduced when key-pecking also produces the S- stimulus. Therefore, an S- stimulus has the potential to be a punishing consequence. This research provides a foundation for the continued investigation of new ethically-viable approaches to punishment. Ongoing research investigates the extent to which an S- stimulus may be used to punish sub-optimal choice strategies.
Stimulus Overselectivity and Effective Reward History
LUCA BLUMHARDT (University of Otago), Brent L. Alsop (University of Otago)
Abstract: Stimulus overselectivity occurs when a response is controlled by specifically selected elements of a compound stimulus. Research suggests that overselected elements may overshadow underselected elements, rather than have more attention allocated towards them, because underselected elements may control responses when overselected elements are later associated with extinction. Presently, there is no mechanism explaining why overselected elements are more salient. We postulate that overselected elements appear more salient because they function more frequently as antecedent stimuli in contingencies in an organism’s learning history. In the present study, 25 undergraduate students were trained to respond to one of two compound stimuli. The colour of each element of the reinforced stimulus was previously associated with differential rewards. We then presented each element of compound stimuli alone, to test whether the overselected element contained the colour previously associated with the highest reward. All elements from the reinforced stimulus were selected equally, irrespective of the reward history associated with each element colour. When arbitrary elements of low or high selectivity were put in extinction, control by these elements usually increased or remained unchanged. Thus, the emergence of previously underselected elements in previous studies may be an artefact of procedural designs, rather than evidence for overshadowing.
Effects of Conflicting Information on Temporal Discrimination
STEPHANIE GOMES-NG (The University of Auckland), Douglas Elliffe (The University of Auckland), Sarah Cowie (The University of Auckland)
Abstract: The environments in which organisms interact are complex, and hence multiple stimuli may come to control behavior (divided stimulus control). What happens when information provided by these stimuli conflicts? The present research investigated the effects of conflicting information on discrimination of elapsed time. Pigeons were trained to associate four stimuli with either a 10-s or a 20-s delay to reinforcement. Following training, non-reinforced single-stimulus and compound-stimulus probe trials were introduced. On single-stimulus and non-conflicting compound-stimulus probe trials, maximum responding occurred close to the usual reinforcement time associated with the stimuli. On conflicting compound-stimulus probe trials, maximum response rates tended to occur at a time in between peak responding on single-stimulus probe trials ('temporal averaging'), replicating previous research using rats. However, statistical analyses indicated that differences in peak response times were not significant, possibly because the delays to reinforcement used in training were similar. Additionally, not all subjects showed evidence of temporal averaging. Nevertheless, these results provide some evidence suggesting that when organisms are presented with two stimuli that provide conflicting information about the time to reinforcement, both stimuli exert joint control over responding.
Stimulus Control in Two Rodent Models of Attention-Deficit/Hyperactivity Disorder
ADAM E. FOX (St. Lawrence University), Sierra Caramia (St. Lawrence University), Molly Haskell (St. Lawrence University), Aerial Ramey (St. Lawrence University), Depika Singha (St. Lawrence University)
Abstract: The spontaneously hypertensive (SHR/NCrl) rat is one of the most widely used models of the combined subtype of Attention-Deficit/Hyperactivity Disorder (ADHD-C). Although often used as its control strain, the Wistar Kyoto (WKY/NCrl) has been proposed as a model of the predominately inattentive subtype of ADHD (ADHD-PI). In Experiment 1 SHR/NCrl, WKY/NCrl, and Wistar (WI; the progenitor strain for the two models) rats were trained on a leftright lever-press sequence in the presence of light discriminative stimuli that signaled the active lever in the sequence. SHR/NCrl and WI accuracy was more disrupted than WKY/NCrl accuracy when light cues were removed or reversed—an indication that behavior of the WKY/NCrl rats may not have come under control of the discriminative light cues. In Experiment 2, all three strains were exposed to a response-initiated fixed-interval (RIFI) 8-s schedule of reinforcement. In RIFI schedules behavior must be timed from a past instance of the target response. Replicating previous work, timing during the FI was equivalent across the three strains, however, latencies to initiate the FI were significantly longer for SHR/NCrl than WKY/NCrl and WI rats. These findings may help determine the efficacy of SHR/NCrl and WKY/NCrl as models of ADHD subtypes in humans.



Back to Top
Modifed by Eddie Soh