|
| Recent Developments in the Study of Mechanisms Underlying Suboptimal Choice |
| Sunday, May 25, 2025 |
| 3:00 PM–3:50 PM |
| Convention Center, Street Level, 151 AB |
| Area: EAB; Domain: Basic Research |
| Chair: Paul Cunningham (University of Minnesota) |
| Abstract: Research in suboptimal choice has shown that some organisms are willing to forgo reinforcement in exchange for signals of reinforcement. Pigeons and starlings consistently prefer an alternative associated with stimuli that reliably signal the presence or absence of food over an alternative that do not reliably signal reinforcement, even if the former has a lower probability of reinforcement than the latter. In this symposium, presenters explore variables linked to theoretical assumptions that contribute to understanding the mechanisms underlying suboptimal choice. The first presentation explored the role of incentive salience of stimuli on conditioned inhibition and found that using tones instead of illuminated keys reduced suboptimal choice but did not have a systematic effect on conditioned inhibition. The second presentation presented the alternatives sequentially instead of simultaneously and evaluated rates of responding when an observing response was required to view the terminal stimuli, showing that suboptimal behavior remained. In the final presentation, a mathematical model for understanding and predicting suboptimal choice based on Delay Reduction Theory is presented and illustrates the importance of temporal parameters in the procedure. Collectively, these studies show variables that influence suboptimal choice that might contribute to elucidate broader behavioral processes governing preference for signals of reinforcement. |
| Instruction Level: Intermediate |
| Keyword(s): Conditioned Reinforcement, Information, Pigeons, Suboptimal Choice |
| |
| Using Tones Reduces Suboptimal Choice in Pigeons: No Evidence of Systematic Effects on Conditioned Inhibition |
| ANGEL M. VILLALOBOS (University of Florida), Cristiano Valerio dos Santos (University of Guadalajara) |
| Abstract: Pigeons prefer an alternative that reliable signals the presence or absence of reinforcement, but with a low probability of reinforcement, over an alternative associated with a higher probability of reinforcement with stimuli that do not reliably signal the reinforcer. This suboptimal choice has not been successfully replicated with rats. It has been suggested that the differences might be due to the incentive salience of the stimuli employed with each species or to a difference in the sensitivity to the stimulus that signals the absence of reinforcement, a presumed conditioned inhibitor. The aim of the study was to evaluate the effects of the incentive salience of stimuli on suboptimal choice and conditioned inhibition in pigeons. Six pigeons chose between two alternatives associated with a 0.2 and a 0.5 probability of reinforcement, respectively. The discriminative stimuli were illuminated keys in one condition and tones in a second condition. Summation tests for conditioned inhibition were conducted for each condition. We found lower suboptimal choice in the tone condition relative to the keys condition. We did not found evidence of a systematic effect on conditioned inhibition. The results are discussed in terms of the enhanced conditioned reinforcement by stimuli attributed with incentive salience. |
| |
| The Paradox of Information: Food-Information Preferences and Observing Responses in Pigeons and Starlings |
| MARCO VASCONCELOS (University of Aveiro), Susana Vieira (University of Aveiro), Armando Machado (University of Aveiro, Portugal) |
| Abstract: Recent research shows that some species behave sub-optimally by "trading" food for information. Suppose animals choose between two options, one Informative and the other Non-informative. If they choose the Informative option, food is delivered in the 20% of trials where an S+ appears, while an S- in the remaining 80% results in no food. If they choose the Non-informative option, one of two stimuli appears, and food is delivered 50% of the time, regardless of the stimulus. Even though the overall probability of food in the Informative Option is 2.5 times lower than in the Non-informative option (20% vs. 50%, respectively), pigeons and starlings strongly prefer the Informative option. In this series of experiments, we investigated whether the suboptimal preference observed in previous studies persisted when choices were presented sequentially rather than simultaneously (Experiment 1) and whether starlings would engage in an observing response to view the terminal stimuli (Experiment 2). Our results showed that suboptimal preferences remained when choices involved either accepting or rejecting a single option presented in isolation. Additionally, we found a high overall rate of observing responses. These findings are discussed in light of current theoretical models. |
| |
| Why Choose Less Reinforcement Instead of More? |
| JEFFREY PISKLAK (University of Alberta), Maggie A. McDevitt (McDaniel College), Marcia Spetch (University of Alberta), Roger Dunn (San Diego State University) |
| Abstract: Nearly half a century ago, an intriguing discovery emerged: Under some conditions, hungry pigeons learn to prefer choices offering significantly less food over those offering more. Since that initial work, considerable research has been directed at understanding the reasons for this behavior which has been variously dubbed suboptimal, maladaptive, or paradoxical because it seems to violate conventional views about the nature of foraging and reward. The behavior has been consistently replicated in birds and appears to parallel the more general literature on choice between smaller sooner and larger later reinforcers. Drawing on insights from a robust theory of choice called Delay Reduction Theory, we demonstrate that suboptimal preference in pigeons can be understood and predicted using well established principles of conditioned reinforcement, as outlined in a new mathematical model known as the Signal for Good News (SiGN) model. Data will be presented to illustrate the importance of temporal parameters in suboptimal choice, in accordance with predictions of the SIGN model. |
|
| |