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.

Search

48th Annual Convention; Boston, MA; 2022

Event Details


Previous Page

 

Poster Session #521
EAB Monday Poster Session: Even-Numbered Posters
Monday, May 30, 2022
2:00 PM–3:00 PM
Exhibit Level; Exhibit Hall A
Chair: Karen M. Lionello-DeNolf (Assumption University)
6. An Application of Machine Learning to Detect the Presence of Challenging Behavior
Area: EAB; Domain: Basic Research
SETH WALKER (Munroe-Meyer Institute), Walker Arce (University of Nebraska Medical Center), Jordan DeBrine (University of Nebraska medical Center, Munroe-Meyer Institute), James Gehringer (University of Nebraska Medical Center), Amanda Zangrillo (University of Nebraska Medical Center, Munroe-Meyer Institute)
Discussant: Karen M. Lionello-DeNolf (Assumption University)
Abstract: In assessing and treating challenging behavior, a significant amount of time and resources are allocated to recording the presence of challenging behavior. Researchers use several different strategies to capture change in behavior over time including permanent product, continuous, and discontinuous methods. One barrier to the use of continuous data collection procedures is that observers must constantly attend to the participant of interest and a second independent observer often simultaneously observes the participant of interest to ensure measurement reliability. One possible solution to resource concerns associated with measurement procedures is to use machine learning models to automate the identification of challenging behavior via video. The primary purpose of this study was to assess the feasibility of using a machine learning model to detect the presence or absence of aggression in the form of hitting. We used a supervised machine learning model trained with a private dataset from one participant consisting of video examples of ‘hitting’ and ‘not-hitting’. To assess the generality of our machine learning model, we tested its performance in identifying novel videos containing hitting. We discuss the implications of our findings and future directions of using machine learning algorithms in behavior analysis.
 
8. Effects of Biologically Neutral Outcomes on a Two-Choice Conditional Discrimination Procedure with Rats
Area: EAB; Domain: Basic Research
MARIO SERRANO (UNIVERSIDAD VERACRUZANA), Pedro Elio Rey Murrieta (Universidad Veracruzana)
Discussant: Karen M. Lionello-DeNolf (Assumption University)
Abstract: The present experiment explored the differential outcomes effect using “biologically neutral outcomes”. Two groups of rats were exposed to a two-choice conditional discrimination procedure. Responses to one of two available levers in the presence of one visual stimulus and responses to the other lever in the presence of a second visual stimulus were reinforced with food pellets. For one group, an auditory stimulus was consistently correlated with reinforcer delivery for responses in one lever but not the other. For the other group, the auditory stimulus was or wasn’t presented with reinforcement after either correct response, randomly. Global percentage of correct responses was similar between groups, but rats exposed to the correlated condition showed a bias for the lever that produced food only. Along with previous experiments, the present results question the differential outcomes effect using either biologically neutral or biologically non-neutral outcomes.
 
10. What-Where-When Remembering in Rats II: Automated Incrementing Non-match-to-sample Procedure
Area: EAB; Domain: Basic Research
SOPHIE LORRAINE PINNEKE (University of North Carolina Wilmington), Hawken V. Hass (University of North Carolina Wilmington), Cassondra Giarrusso (University of North Carolina Wilmington), Elijah Richardson (University of North Carolina Wilmington), Spencer Bruce (University of North Carolina Wilmington), Madeleine Mason (University of North Carolina - Wilmington ), Mark Galizio (University of North Carolina Wilmington), Katherine Ely Bruce (University of North Carolina Wilmington)
Discussant: Karen M. Lionello-DeNolf (Assumption University)
Abstract: The Odor Span Task has been used to examine control by multiple stimulus properties (e.g., what-where-when components of an event) in rats in an arena setting. The present study extended this research by assessing what-where-when remembering in two distinct contexts using an automated, go-no-go, incrementing non-match-to-sample procedure in five rats. Reinforcement was delivered only for context-novel odor stimuli in each context. Phase 1 included one transition between contexts to measure rats’ discrimination of which odor was presented in which context. Phase 2 increased the number of transitions during the session to two and three context changes. In Phase 3, probe trials were added to control for recent familiarity with odor stimuli. Rats learned to respond to context-novel scents at above chance levels with one, two, or three context transitions during Phases 1 and 2. However, in Phase 3, performance on probe trials comparing item-in-context to familiarity-based responding was indiscriminate, and modifications to the training procedure are ongoing. Developing an automated procedure to study episodic-like remembering in rats would be useful to eliminate possible sources of human error, minimize handling of subjects between trials, control confounding variables, and expand further research on therapeutic approaches for different memory disorders in humans.
 
12. The role of conditioned reinforcer and discriminative stimulus in a token reinforcement system
Area: EAB; Domain: Basic Research
RICARDO SILVESTRE CAMPOS RIVERA (University of Guadalajara), Cristiano Valerio dos Santos (Universidad de Guadalajara)
Discussant: Karen M. Lionello-DeNolf (Assumption University)
Abstract: On the several functions of the tokens, the roles of conditioned reinforcer and discriminative stimulus have received more attention. Five pigeons were exposed to a second order schedules in a two-component multiple schedule: master and yoked component. In Condition A, tokens were delivered in the master component according to a Variable Interval 15 s (VI 15) and the component ended when four tokens were obtained. In the yoked condition, the tokens were programmed according to the time that they were obtained in the previous master component. In Condition B, the master component was a conjoint VI 30 and a Variable Time 30 s that ended after the delivery of four tokens. The four tokens of the yoked component were response-dependent and programmed according to the time they were obtained in the previous master component. Condition C was similar to A with the differences that tokens were extinguished. Tokens or food were exchanged at end of each component. Results showed an increase in the response rate as function of obtained tokens which suggest a discriminative function and also a relative decrease in response rate of the free-token component but not in the yoked component which supports the idea of conditioned reinforcer.
 
14. What-Where-When Remembering in Rats I: Variations on the Odor Span Task
Area: EAB; Domain: Basic Research
HAWKEN V. HASS (University of North Carolina Wilmington), Emily Burrell (University of North Carolina Wilmington), Rebeca Barba (University of North Carolina Wilmington), Dylan Marshall (University of North Carolina Wilmington), Amanda Burke (University of North Carolina Wilmington), Mark Galizio (University of North Carolina Wilmington), Katherine Ely Bruce (University of North Carolina Wilmington)
Discussant: Karen M. Lionello-DeNolf (Assumption University)
Abstract: Behavior can come under the control of multiple stimulus properties (e.g., what, where, and when components of an event); this has been termed “episodic-like” remembering. We used two procedural variations of the Odor Span Task (OST) to examine episodic-like remembering in rats. In the OST, selection of a session-novel odor results in reinforcement, but odors are not reinforced on subsequent presentations. In the Context variation, three rats were trained on the OST in two distinct apparatuses each day; selection of context-novel odors resulted in reinforcement. In-context probe trials after two context transitions indicated above chance performance even when controlling for familiarity cues. In the Flavor variation, one odor was designated each day as a replenishing odor, while the other OST odors did not replenish after their first presentations. Selection of the replenishing odor was reinforced with a berry-flavored pellet, while other odors were baited with sucrose pellets. Four rats learned which odor was associated with berry (what) each day (when). These procedures show promise as models for testing episodic-like remembering in laboratory rats. Such animal models have important clinical implications for developing treatments for disorders characterized by loss of episodic remembering, such as Alzheimer’s disease.
 
16. Repeated Renewal During Dense and Lean Schedules of Differential Alternative Reinforcement: A Human Operant Investigation
Area: EAB; Domain: Basic Research
RYAN KIMBALL (University of Saint Joseph), Lindsay Day (University of Saint Joseph), Emily Ferris (University of Nebraska Medical Center; Munroe Meyer Institute), John Silveira Jr. (University of Saint Joseph), Rebecca Karis (University of Saint Joseph)
Discussant: Karen M. Lionello-DeNolf (Assumption University)
Abstract:

Operant renewal is a form of treatment relapse that occurs when a previously suppressed response reemerges due to a change in context. Previous research has demonstrated that renewal of target responding may occur despite the availability of differential reinforcement for an alternative response (DRA). Nevertheless, the current literature on renewal presents mixed findings regarding the impact of dense and lean schedules of DRA on the magnitude of renewal. In addition, little is known about the impact of repeated renewal tests. We used a translational approach to study the effects of dense and lean schedules of DRA during repeated renewal tests with undergraduate college students and a simulated computer task. All participants experienced two, three-phase ABA renewal arrangements. In the dense and lean renewal arrangements, we differentially reinforced alternative behavior in Context B and the renewal test in Context A on a VI 3-s or a VI 12-s schedule, respectively. Overall renewal effects were small. Further, the data suggest that although renewal is possible in both arrangements, a slightly higher magnitude of renewal may be more likely with a lean schedule of reinforcement versus a dense schedule. However, the data also suggest that repeated tests may decrease the magnitude of renewal.

 
18. An Examination of Within-Session Operant Response Patterns in Mice: Satiation or Habituation?
Area: EAB; Domain: Basic Research
NICHOLAS L VITALE (University of Nevada, Reno), Matthew Lewon (University of Nevada, Reno)
Discussant: Karen M. Lionello-DeNolf (Assumption University)
Abstract: Previous research has suggested that systematic decrements in operant response rates often observed within experimental sessions are consistent with habituation to the repeated presentation of reinforcers rather than other factors such as fatigue or satiety. One way to test for habituation is to conduct a test for dishabituation, a phenomenon in which a temporary recovery of responding is observed following some brief stimulus change. Dishabituation of operant responding has been demonstrated on several occasions in the literature; however, studies with non-humans have thus far been limited to those using rats and pigeons as subjects. Therefore, the purpose of the current study was to attempt to replicate these findings with mice. During baseline, two groups of mice nose-poked for sweetened condensed milk on either a fixed-ratio 4 or variable-interval 15s schedule of reinforcement. For each group, baseline sessions were then alternated with two test conditions and a control condition until each was conducted three times. Test conditions included a 5s tone or flashing of the houselight presented halfway through the session. Control conditions were identical to baseline. Results indicated that dishabituation was not observed for either group. Several methodological considerations for further study of operant dishabituation in mice will be discussed.
 
20. Examining Underlying Behavioral Mechanisms and Effectiveness of Cues Paired with Alternative Reinforcement in Mitigating Resurgence
Area: EAB; Domain: Basic Research
CARLA N MARTINEZ-PEREZ (Auburn University), Toshikazu Kuroda (Advanced Telecommunications Research Institute International), Carolyn Ritchey (Auburn University), Christopher Podlesnik (Auburn University)
Discussant: Darlene E. Crone-Todd (Salem State University)
Abstract:

Problem behaviors such as aggression and self-injury are prevalent among children diagnosed with autism spectrum disorder (ASD). Fortunately, treatments such as differential reinforcement of alternative behavior (DRA) are efficacious in reducing problem behaviors among children with ASD. However, changes in treatment conditions such as treatment integrity errors can result in resurgence of problem behavior. Surprisingly, there is no systematic research geared toward developing effective techniques to mitigate resurgence during DRA treatments. While previous research has shown that cues paired with DRA treatments (hereafter treatment cues) can mitigate relapse when compared to the absence of such cues, there been no research dedicated toward either (1) understanding the behavioral mechanisms underlying treatment cues or (2) enhancing the effectiveness of cues for DRA treatments in mitigating resurgence. We addressed each of these questions in a series of translational experiments with MTurk participants after demonstrating that treatment cues effectively mitigated resurgence in that population compared to the absence of cues. Results of the present research will contribute to our understanding of behavioral mechanisms contributing to relapse and the development of more effective DRA treatments.

 
22. The Effects of Contingency Descriptions on Delay Discounting
Area: EAB; Domain: Applied Research
TAYLOR BAKALAR (New England Center for Children; Western New England University), Jason C. Bourret (New England Center for Children), Stefanie Upshaw (The New England Center for Children & Western New England University)
Discussant: Darlene E. Crone-Todd (Salem State University)
Abstract: Delay discounting preparations have been used across species to assess the degree to which delays devalue consequences, where the subjective value of a reward decreases with increasing time. Among studies that use these preparations with humans, none assess indifference points in the absence of contingency descriptions (i.e., no prior research with humans has examined discounting of consequences without trial-by-trial contingencies being described to participants). In the current study, two remote experimental arrangements were implemented to assess the effect of these contingency descriptions on discounting in 15 typically developing adults. In Experiment 1, 10 adults selected a delayed or immediate contingency by sending a letter through text to the experimenter, who would send monetary reinforcement through an online application. Delays assessed included 30 s, 15 s, 5 s, and no delay. Contingency description and reinforcer presentation differed across three conditions: reinforcement in the absence of verbal contingency descriptions, reinforcement with verbal contingency descriptions, and verbal contingency descriptions with no reinforcement delivery. Experiment 2 followed the same general procedures, with an added test to ensure that behavior was sensitive to the differential magnitudes manipulated. Across both studies, participants discounted more steeply in the reinforcers only condition relative to conditions including contingency descriptions.
 
24. Molecular Feedback to Signal Reduction of an Avoidance Session: Enhanced Molar Control or Conditioned Reinforcement?
Area: EAB; Domain: Basic Research
JEREMY SAUL LANGFORD (West Virginia University), Cory Toegel (Northern Michigan University), Catherine Williams (Marcus Autism Center), Michael Perone (West Virginia University)
Discussant: Darlene E. Crone-Todd (Salem State University)
Abstract: Identifying conditions under which behavior is sensitive to molar versus molecular contingencies remains an important topic of study. The present study sought to improve control by a molar contingency by adding molecular feedback. Four rats pressed two levers associated with a single Sidman avoidance schedule; a press on either lever avoided shock. In addition, in some conditions presses on one of the levers reduced the total session duration by 1 min. Moving the session-reduction contingency between the left and right levers had no reliable effect on response allocation. In other conditions, a molecular consequence was added to the session-reduction lever: Each press produced a feedback stimulus. The feedback tended to increase responding on the session-reduction lever. In a control condition in which the feedback was provided without the session-reduction contingency, however, responding still tracked the lever that produced the feedback. Tracking of the feedback stimulus, independent of the session-reduction contingency, suggests that the feedback may function as a conditioned reinforcer. Additional research is underway to address this potential account. These findings illustrate the elusive nature of control by molar contingencies, even when accompanied by molecular feedback.
 
28. Molar perspective of the Experiential Discounting Task Between Different Rewards
Area: EAB; Domain: Basic Research
SILVIA MORALES-CHAINE (National Autonomous University of Mexico), Carlos Adrián Jandete (Autonomous National University of Mexico)
Discussant: Darlene E. Crone-Todd (Salem State University)
Abstract:

Delay discounting has been studied through animal models and psychophysical adjustment procedures. Many procedures have been developed to assess delay discounting in humans and many of these procedures using diferente rewards. The Experiential Discounting Task (EDT)was developed to assess human delay discounting using primary, secundary and general rewards. In the present study we examined the delay discount and the impulsivity measured through the area under the curve using different rewards. Construct validity was evaluated by comparing the EDT giving a reward with videogame time with an EDT with food reward, time playing videogame and a standard delay discounting task. We stablished 4 evaluations keeping the same delays. Both EDT had poor test–retest reliability and discounting rates obtained with those task were uncorrelated with the other rewards. R square above 75% in the individual. We also measured the value assigned to the reinforcer with an visual analog scale and found the same value in each medition. These findings may suggest that there are not measured variables beyond the stablishment operations and demolition operations. Also the EDT measures point out that a different construct of DT is measured by traditional delay discounting tasks with different rewards

 
30. Behavioral Measures of Gain-Loss Asymmetry in Decision-Making and Choice with Food
Area: EAB; Domain: Basic Research
MARCIA M. VENTURA (Brigham Young University), Blake Hansen (Brigham Young University), Rebecca Lundwall (Brigham Young University ), Harold Miller, Jr. (Emeritus Brigham Young University )
Discussant: Darlene E. Crone-Todd (Salem State University)
Abstract: The asymmetrically greater effect of losses on behavior, when compared to gains of the same objective value, is a key concept in behavioral economics. This phenomenon, known as loss aversion, results in a preference for avoiding loss rather than pursuing gains, regardless of reinforcement allocation. Cognitive and behavioral methods are typically used to investigate loss aversion in monetary and other quantitative domains. Instead, we examined gain-loss asymmetry in decision-making and choice with gains and losses of real food. We used a computer game to directly measure behavioral allocation in 6-ply interdependent concurrent VI VI schedules of reinforcement (gain = +1 food point) and punishment (loss = -1 food point). Points were exchanged for food at the conclusion of daily experimental sessions. We used the generalized matching law to derive sensitivity and bias parameters. We calculated gain-loss differentials using a pairwise contrast of the bias parameters from gains-only conditions and gains-plus-punishment conditions. Nine college students (4 females), aged 18-25, completed 8 sessions. As predicted, we found an asymmetrically greater effect of losses relative to gains of food. We also found that gain-loss asymmetry ratios were substantially higher than those typically reported in experiments involving hypothetical or actual amounts of money.
 
 

BACK TO THE TOP

 

Back to Top
Modifed by Eddie Soh
DONATE