Association for Behavior Analysis International

Experimental behavior analysis has produced a robust literature on the role of bias in matching and choice. Despite its beginnings in the laboratory, this work has extended far beyond simple choice models and operant responses to address how bias impacts social responding. This symposium will present contemporary research on the assessment of implicit bias, translational models of choice, and the ways in which bias can impact the choices researchers make within the laboratory.

Board certified behavior analysts; licensed psychologists; graduate students. 

The study of implicit bias in behavior analysis has been dominated by one particular method, the implicit relational assessment procedure (IRAP). The IRAP could be considered quite unusual as a method for studying implicit bias because it targets verbal relations as defined within relational frame theory (RFT). In contrast, implicit bias in the non-behavior-analytic “mainstream” literature is often interpreted as reflecting the strength of associative links in a mental realm (e.g., a memory store). Despite this conceptual difference, research on implicit bias using the IRAP could be seen as relatively successful, at least in terms of number of published studies and the results of a meta-analysis of IRAP studies. On balance, until relatively recently IRAP research tended to focus on the method as a measure of implicit bias without conducting experimental analyses of the multiple variables, from an RFT perspective, that are brought into play when participants complete an IRAP. Conducting these more recent experimental analyses has served to produce an increasingly sophisticated and complex understanding of exactly what so called “implicit bias” involves from an RFT and a behavior-analytic perspective. The current paper will provide an overview of this research story.

Delay discounting measures a facet of impulsivity and is related to various socially significant behaviors. Researchers suggest that altruism and impulsivity arise from the same underlying mechanism; thus, individuals are less altruistic towards people that are different from themselves (e.g., a different race or ethnicity). However, researchers have yet to analyze the relation between delay discounting and implicit and explicit racial bias. In the present study, participants will complete a delay discounting procedure and several Likert scale surveys: the Implicit Relational Assessment Procedure (IRAP), the Color-Blind Racial Attitudes Scale (CoBRAS), and the Scale of Ethnocultural Empathy (SEE). The delay discounting task will assess discounting for two delayed magnitudes (i.e., $100 & $1000) using an adjusting amount procedure. The IRAP measures implicit racial bias by having participants categorize stimuli based on either pro-Latino stereotypes or pro-White stereotypes; the difference in the length of time required to categorize stimuli that are consistent or inconsistent with the stereotypes measures bias. To examine explicit racial bias, we will use the total scores on the CoBRAS and the SEE. We expect participants who steeply discount delayed rewards will be more implicitly and explicitly racially biased toward members of a different race or ethnicity.

Historically, women have experienced bias in science. This bias has affected women as participants in research, as well as women as conductors of research. Although little recognized, another form of bias extends to female laboratory animals. Female laboratory animals, particularly rodents, have long been regarded as more variable in their behavior and other dependent measures and therefore were excluded from experiments. I will describe the inclusion of female laboratory animals in the Journal of the Experimental Analysis of Behavior. I will describe trends over time in the inclusion of female laboratory animals for different types of animals (e.g., pigeons and rats). For studies involving both sexes and in which animals are identified by sex, I will examine the level of variability in the behavior of male and female animals. Finally, I will review published data describing the variability in a wide variety of dependent measures for both male and female laboratory animals. These data show that female laboratory animals have been excluded from research based on bias rather than fact. The exclusion of female laboratory animals precludes learning about genuine sex differences with important health implications and is no longer allowed in NIH-funded research.