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 #434
CE Offered: BACB
Quantitative Modeling of Choice Behavior and Extensions to Caregivers of Children With Challenging Behavior
Monday, May 29, 2017
10:00 AM–11:50 AM
Hyatt Regency, Centennial Ballroom B/C
Area: EAB/CBM; Domain: Translational
Chair: Kenneth Shamlian (University of Rochester School of Medicine )
Discussant: Amy Odum (Utah State University)
CE Instructor: Kenneth Shamlian, Psy.D.

Experimental behavioral economics is a combination of behavioral analysis and economic concepts, principles, and measures to model choice behavior at the individual level. Of particular interest is the bias for individuals to favor more immediate, but smaller rewards over larger rewards that would come later (i.e. delay discounting). This work has resulted in a wealth of research showing that discounting tasks and predictive mathematical models can quantify sensitivity to delayed outcomes. The focus for this symposium will be translational in nature and discuss related research findings for: (a) foundations and methods for understanding mathematical modeling of choice behavior, (b) considerations when comparing measures of discounting across studies, (c) simulated parent and teacher discounting of delayed treatment effects for children with problem behavior, and (c) patterns of caregiver discounting of delayed treatment effects in an outpatient setting providing manualized caregiver training and individualized, behavioral-caregiver therapy for children with challenging behavior.

Instruction Level: Intermediate
Keyword(s): Challenging Behavior, Delay Discounting, Parent Choice, Quantitative Models
A Qualitative Comparison of Quantitative Models of Economic Demand
(Basic Research)
LINDSAY LLOVERAS (New England Center for Children), Jason C. Bourret (New England Center for Children), Joshua Jackson (New England Center for Children)
Abstract: Mathematical models are quantitative descriptions of functional relations. They can be evaluated in a number of ways. Quantitative evaluations can include a comparison of percentages of variance accounted for, sums of squared residuals, and AIC or F-test values. In addition, they can be evaluated in terms of the distributions of residuals around fit lines. While this may sound thoroughly complex and complicated, it actually is not. In this talk, we compare fits of the Hursh (1988) linear-elasticity model and the Hursh and Silberberg (2008) exponential-demand model. This talk is designed for people who would like to learn more about mathematical modeling in order to develop a sufficient repertoire to appropriately consume the quantitative analysis literature. Interobserver agreement was collected for 19.6% of data sets; mean agreement for price and mean agreement for consumption were 97.4% and 98.8% respectively.
Effects of Standardizing the k Parameter in the Exponential Model of Demand
(Basic Research)
JOSHUA JACKSON (New England Center for Children), Jason C. Bourret (New England Center for Children), Lesleigh Ann Craddock (New England Center for Children)
Abstract: The exponential demand model is used to quantitatively describe the functional relation between the consumption of a particular commodity and its price. The value of the k parameter of the exponential demand model directly impacts the values of both the free parameters of the model, Q0 and α, as well as the model’s goodness of fit. Thus, comparison of a across commodities requires the use of a common k in the curve fits for both. This poses a challenge for across-study comparisons in which different k values are used. We examined the effects of using study-specific, differing k values versus a single k value in fitting the exponential demand model to 269 previously published sets of demand data. Changing from study-specific to a single k value across studies resulted in a change in the ordinal rank of commodities’ reciprocal a values in 43% of studies. The use of a common k across studies did not have a meaningful effect on the goodness of fit of the model. Taken together, researchers should pay particular attention to any changes in the ordinal rank of commodities when fitting demand data with a common k for comparison.
The Impact of Delays to Treatment Success on Perceptions of Interventions for Problem Behavior
(Basic Research)
MINDY CHRISTINE SCHEITHAUER (Emory University; Marcus Autism Center), Christina Simmons (University of Georgia; Munroe-Meyer Institute; Uni), Nathan Call (Emory University; Marcus Autism Center), Jamieson Ekstrom (Marcus Autism Center)
Abstract: The concept of delay discounting may have interesting implications for behavioral interventions targeting problem behavior. Specifically, for treatments that include extinction, a caregiver often has a choice of reinforcing the problem behavior now (resulting in immediate cessation) or implementing extinction and working through the problem (potentially resulting in continued problem behavior now, but decreased problem behavior in the future). The purpose of this study was to examine discounting of delayed treatment effects using a translation model. A group of college students (N = 113) read vignettes depicting problem behavior and selected either an immediate but small treatment effect (e.g., 20% of treatment goals met immediately) or delayed but larger treatment effect (e.g., 100% of treatment goals met in 4 weeks). In addition, participants were divided into four groups, each of whom read different vignettes that varied on two dimensions: type of caregiver (teacher or parent) and severity of problem behavior (high or low). Results show that discounting of treatment effects matched well with the discounting model and no significant differences were found across groups, suggesting that discounting occurred despite the type of caregiver or severity of problem behavior included in the vignette.
Temporal Discounting of Delayed Treatment Effects in Caregiver-Mediated Therapy
(Service Delivery)
KENNETH SHAMLIAN (University of Rochester School of Medicine), Shawn Patrick Gilroy (National University of Ireland, Galway)
Abstract: Research has established that people have a bias towards selecting an option that provides some desired amount of a commodity more quickly regardless if waiting would produce a more favorable outcome (i.e. discounting the value of a better choice). Caregivers of children with challenging behavior may be more likely to show patterns of responding in a way that alleviates stress or harm more quickly; even when better, long-term outcomes are available. Currently, only one study has examined parents’ perceptions of treatments and the impact of delayed outcomes typically observed during treatment for a child’s problem behavior. The purpose of this study was to: (a) determine the extent to which caregiver’s choice making for delayed treatment outcomes fit established discounting models, (b) assess whether a computer-based delay discounting task provides an efficient method for assessing caregiver bias against delayed treatment outcomes, and (c) determine the extent which caregiver discounting patterns correlate with clinical measures and outcomes. Caregivers were divided by their respective caregiver-mediated therapy modality (manualized parent training vs. individual behavior therapy) and provided a tablet-based adjusting amount task for selecting a sooner, smaller reduction in problem behavior (e.g. 15% reduction immediately) versus a larger, later reduction in problem behavior (e.g. 100% in 4 weeks) across seven different delays in time. Initial results suggest that caregivers' patterns of discounting match established models of temporal discounting and can be conducted within the course of outpatient treatment efficiently (total task time; M= 5 min) . Additionally, measures of treatment adherence, barriers to treatment adherence, and population demographic characteristics are discussed in comparison to discounting patterns.



Back to Top
Modifed by Eddie Soh