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.


31st Annual Convention; Chicago, IL; 2005

Event Details

Previous Page


Symposium #376
Int'l Symposium - The Implicit Association Test (IAT): Behavior Analytic Research Applications and Interpretations
Monday, May 30, 2005
2:30 PM–3:50 PM
Waldorf (3rd floor)
Area: TPC; Domain: Theory
Chair: Maria R. Ruiz (Rollins College)
CE Instructor: Maria R. Ruiz, Ph.D.
Abstract: The Implicit Association Test (I.A.T.) developed by Anthony Greenwald is said to measure unconscious cognitions that influence attitudes and preferences towards a wide range of social phenomena including race, politics and views on sexual orientation.Interestingly, the Implicit Association Test is based on a methodology that can be easily incorporated in behavior analytic research. Specifically, the I.A.T. technique bears a striking similarity to behavioral methods of attitude and behavior assessment that rely on the concept of stimulus equivalence and derived relational responding. Our symposium provides an overview of the I.A.T and outlines its relevance to behavioral research in the social sciences. We will present a behavioral model of the IAT in terms of derived stimulus relations and data supporting the behavioral account will be outlined. Data illustrating the operant nature of results obtained with I.A.T preparations that are sufficient to account for processes described as “attitudes” and “racial prejudice” will be presented. Finally, an update on a behavioral model of the I.A.T as a screening test for sex offenders will be described.
The Implicit Association Test (IAT): Conceptual Analysis of “Unconscious Cognitions” as Explicit Verbal Histories
MARIA R. RUIZ (Rollins College), Bryan T. Roche (National University of Ireland, Maynooth)
Abstract: The Implicit Association Test developed by Anthony Geenwald exposes participants to two-element compound stimuli consisting of a face (Caucasian or African American) and either a positive or negative word (good/bad), and asks them to categorize novel stimuli (faces and positive/negative words) accordingly. A significant proportion of participants respond with greater accuracy in the categorization of novel stimuli when the positive word is combined with the Caucasian face. This effect (which we are referring to as the IAT effect) is said to illustrate racial preference (read racial bias). Greenwald suggests that the categorization errors are evidence of implicit associations and unconscious cognitions. Our research presents an alternative behavioral model from which to consider the problem based on derived stimulus relations. We suggest that categorization “errors” in the IAT are likely related to pre-experimental histories of relational responding prevalent in the participants ’verbal community.
Implicit Associations: An Experimental Analysis of Explicit Learning Histories: An Update
KEVIN M. MIRAGLIA (Rollins College), Maria R. Ruiz (Rollins College), Bryan T. Roche (National University of Ireland, Maynooth)
Abstract: We will present an experimental analysis of explicit learning histories sufficient to generate the IAT effect equivalent to that produced in Anthony Greenwald’s Implicit Association Test. After mastering arbitrary equivalence relations involving nonsense syllables participants are exposed to the Implicit Association Test. Specifically, four sets of individually acquired 3-member (nonsense syllables) equivalence classes associated with four colors are linked to form two separate superodinate (6-member) classes. These conditions model the race-word combinations in the Greenwald test. Categorizations of novel stimuli with derived color functions replicate Greenwald’s IAT effect. Subjects categorize novel stimuli with derived color functions “Within” superordinate classes (corresponding to pre-experimental culturally reinforced relational responding) with great accuracy. Participants consistently commit errors when categorizing these “Across” superordinate classes (corresponding to equivalence relations typically not selected by the verbal community). Implications of this research for understanding ‘implicit associations’ as explicit behavioral processes will be discussed.
A Derived Relations Approach to Screening Tests for Sex-Offenders: An Update
BRYAN T. ROCHE (National University of Ireland, Maynooth), Maria R. Ruiz (Rollins College), Martina O'Riordan (National University of Ireland, Cork), Ethel Quayle (National University of Ireland, Cork), Maxwell Taylor (National University of Ireland, Cork)
Abstract: Behavior analysts working in the fields of stimulus equivalence and Relational Frame Theory have already provided the conceptual and technical foundation for the development of screening tests for a range of normal and pathological behavior patterns. These tests are based on the broad idea that individuals' verbal history will impact on their performance in laboratory-based training in derived relations in predictable ways. These procedures come strikingly close to the core processes that appear to be at work in the Implicit Association Test (I.A.T). The current study will outline how the concept of derived stimulus relations has been used to develop a screening procedure for the assessment of pedophiles. More specifically, an I.A.T-type procedure was used in a single-blind study to examine differences in relational responding between contact sex-offenders against children, internet paedophiles, sex offenders against the adult, non sex-offender criminals, and a random sample from the general population. The results suggest that while contact sex offenders are distinguishable from the normal population using these tests, convicted users of child pornography are not. The utility of the IAT as a behavioral assessment tool and the theoretical implications of these findings will be discussed.
The IAT: A Measure of Relational History or Attitudes?
AMANDA GAVIN (National University of Ireland, Maynooth), Bryan T. Roche (National University of Ireland, Maynooth), Maria R. Ruiz (Rollins College), Carmel Gormley (National University of Ireland, Maynooth)
Abstract: In Experiment 1, Subjects were exposed to a word-picture association training phase using a respondent conditioning preparation. Specifically, two nonsense syllables were paired with sexual and disgusting images, respectively. The nonsense syllable paired with the sexual images was blue in color, while that paired with the disgusting images was red in color. Subjects were then exposed to an equivalence training procedure which led to the formation of two three-member equivalence relations, each containing one of the two nonsense syllables. Subjects were then exposed to an IAT-type test consisting of red, blue, sexual, and disgusting images. Subjects were also exposed to a more complex IAT-type test consisting of sexual and disgusting images, and all members of the trained equivalence relations. Results suggest that the IAT is sensitive to the respondent conditioning history employed in phase one and the verbal relations established by the experimenters. These findings suggest that the IAT may not be a reliable measure of attitudes but more correctly provides an indication of subjects' relational histories that may or not be indicative of attitudes.



Back to Top
Modifed by Eddie Soh