Association for Behavior Analysis International

In this symposium, we will explore multiple layers of a nested model of behavior that impacts earth’s climate. The model expands from solutions at the level of the individual (i.e., individualistic responsibility) to challenges that persist at the level of small (i.e., interrelated dynamic contingencies) and large (i.e., social policies) groups. First, we describe an experimental program to reduce individual carbon emission through a functional assessment of green behavior and an embedded shaping procedure. Second, we will discuss a series of basic experiments that model resource depletion as a function of competition that influences contingencies at the level of the individual and small groups. Third, we will describe how policies at a social level can impose constraints of collective behavior - but that preference for these policies can be successfully modelled within a delay discounting paradigm. Finally, we will move beyond discussing individuals’ context-specific behavior to propose a mathematical model that profiles individuals’ environmental choices across multiple circumstances and domains.

Behavior Analysts

In recent years, concerns relating to global warming and the need for reducing carbon emissions has increased. An effective approach for reducing overall carbon emissions is to increase sustainability related behaviors. While such an approach affords this utility, an underlying factor that potentially limits the extent to which individuals engage in sustainable behavior is limited knowledge or information as to what specific behaviors are considered to be sustainable. As such, there is a need for a methodology to identify deficits as they relate to sustainability behavior. The current symposium discusses the development of an assessment designed to provide a measure of an individual’s sustainability behavior. In addition, researchers sought to evaluate the validity of the assessment by determining the extent to which assessment scores were related to carbon output. Scores for the sustainability assessment as well as carbon footprint measures were collected, and a Pearson product-moment correlation coefficient was obtained between the two measures. The results showed a moderate, negative correlation between scores on the sustainability assessment and carbon footprint measures. These findings suggest the sustainability assessment is a valid tool which has good correspondence with other sustainability measures, and it can be used to identify sustainability related behavior deficits.

Policies provide shared social and economic contingencies that can influence the behavior of large groups, representing the outermost layer of our nested model. We will discuss data collected within multiple delay discounting tasks that have been adjusted to capture policy preferences and willingness to forego high emission commodities to delay the climate point of no return. Results suggest that participants discount climate change similar to catastrophic monetary losses, and that policy manipulations (group versus individual contingencies) can influence willingness to forego emission commodities. Behavior analysts may therefore play a role by quantitatively evaluating preference for policies that target high emission behavior. Research collected during and throughout COVID-19 will also be reviewed as an approximate model to the climate crisis showing that perceived probability of a catastrophic outcome and grouped versus individual contingencies can have a considerable impact on participants’ willingness to alter or adjust behavior to delay or avoid future hardship. Finally, the talk will conclude with an analysis of how the policy layer may conceptually interact with lower layers of the nested model.

Traditionally, discounting data is analyzed between-subjects in the form of indifference points. Separate curves are fit for each combination of conditions, and a curve-fitting parameter, k, is compared across conditions with a t-test or ANOVA. However, it is also possible to analyze this data in a repeated-measures analysis that treats responses as individually related cases. This talk will compare and contrast approaches using environmental discounting data and discuss how a multi-level approach allows researchers to model contingencies that occur on a geographic or partisan level, while preserving and studying relationships that unfold on the level of the individual. The talk will be practically-oriented, with R-markdown text provided as supplemental material.