A Cognitive Model of the Implicit Associations Test

点击翻译按钮获取中文摘要

The study of implicit, automatic attitudes has been a dominant theme in social psychology over the last two decades. While a great deal is known about the causes and consequences of such attitudes, very little is known about the specific cognitive processes underlying implicit attitude judgments. The major aim of this research is to model the cognitive mechanisms underlying one of the most widely used paradigms to test implicit attitudes: the implicit associations test (IAT: Greenwald, McGhee, & Schwartz, 1998). Since its development six years ago, there have been over 100 articles published using the IAT to test a variety of implicit attitudes. Despite its popularity, the meaning of IAT scores remains controversial. Little is known about the cognitive processes driving implicit attitudes, and even less is known about what exactly the IAT in particular assesses. In order to map the cognitive processes underlying the IAT, this project will employ a model that makes explicit assumptions about processing, and explicit quantitative predictions: the diffusion model (e.g., Ratcliff, 1978; Ratcliff & Rouder, 1998). The diffusion model is a model of decision making in tasks wherein participants make fast binary choices (like in the IAT). Explicit models of fast decision making can reveal a number of interesting alternative explanations for decision processes that could illuminate our understanding of tasks that intend to measure implicit attitudes. In a series of IAT studies, different factors of the IAT will be manipulated in order to test different possible cognitive processes underlying IAT effects. To do this, a standard IAT procedure will be employed, and various factors within this paradigm will be manipulated to test the model under different conditions. By applying a quantitative modeling tool to the IAT, this research can better reveal the cognitive processes that are at play when a person takes an IAT. In so doing, three goals can be attained: (1) to advance our field's understanding of the cognitive processes used when participating in the IAT; (2) to use the diffusion model to account for IAT data, and to test different versions of the model that are anchored in theories about the locus of IAT effects; and (3) to establish a data set of different parametric manipulations in the IAT so that future researchers can use the IAT more strategically depending on the cognitive processes they wish to explore. By effectively modeling the IAT, scientists would be better able to use this popular methodology more appropriately and effectively. This is especially urgent given that more and more scientists are interested in exploring intervention applications of the IAT, such as in understanding attitudes related to psychological dysfunctions, consumer behavior, and so on. A better understanding of unconscious contributions to attitudes and behavior could provide greater insight into important social issues, such as attitudes toward racial discrimination.