Mahzarin Rustum Banaji, PhD - US grants

In ongoing research concerning implicit social cognition we have developed the view that important components of the cognitive processes that underlie social actions are (a) ordinarily hidden from the actor's awareness or ability to control and, as a consequence, (b) beyond the possibility of effective investigation using methods that are limited to self-report measures. Our research has successfully employed several indirect measures to reveal the operation of implicit stereotypes. The first goal of this project is to extend these investigations of implicit social cognition to the domains of the social-cognitive constructs of implicit attitudes and implicit self-esteem. Existing measures of unconscious social cognition (most prominently semantic priming measures) have some notable limitations. Although these measures have been used effectively at laboratories in which they were developed, they (a) have not readily been exported to other laboratories and (b) have not yielded large enough effect sizes to succeed as individual difference measures. Our previous NSF-funded research has yielded a new procedure (the Implicit Association Test, IAT) that has been shown to provide a reliable, efficient, and sensitive measure of individual differences in implicit social cognition. The second goal of the current project is to document the basic properties of the IAT method and to develop its use as a core technique for investigating implicit social cognition. The research will yield three types of desirable products: (a) new measures that should be broadly useful in psychological investigations of social behavior, (b) new theoretical insights into the involvement of unconscious cognition in social behavior, and (c)applications in important settings, such as workplace and school, in which unrecognized (i.e., implicit) stereotypes and prejudices interfere with desirable or productive behavior.

For the past eight years, the PIs (M. R. Banaji and A.G. Greenwald) have collaborated on investigations if implicit social cognition. The main goals of the research program have been to demonstrate the operation of implicit social cognition, and to champion the view that theoretical, methodological, and application gains can result from investigations of processes that are ordinarily hidden from awareness. Past work focused on the stereotype domain, with published articles documenting the use of a variety of methods to reveal the operation of implicit stereotypes (Banaji and Blair, 1996; Banaji and Greenwald, 1995; Banaji and Hardin, 1996; Banaji, Hardin, and Rothman, 1993). A first goal of the proposed research is to extend investigations of implicit social cognition to the attitude and self-esteem domains identified by Greenwald and Banaji (1995). Greater attention to the constructs of attitude and self-esteem allows the development of a unified theory of implicit social cognition. Existing measures of automatic belief and attitude (especially ones based on cognitive priming) have been used effectively at some laboratories including our own, but have not been easy to transport across laboratories. It is also the case that existing methods (consisting of variations of priming techniques) have not yielded large effect sizes. Recently published work has produced a new measure (the Implicit Association Test, IAT; Greenwald, McGhee, and Schwartz, 1998) that forms the foundation of the present proposal. Several experiments conducted with the measure indicate that it can provide a reliable, efficient, and sensitive measure of individual differences in implicit social cognition. A second goal of the proposed research is to realize the potential of the IAT, method and to further develop it as a core technique for investigating implicit social cognition. This proposal consists of experiments that will further establish the generality and validity of the technique followed by three sections with a theoretical focus on prejudice and stigma, self-esteem and self-concept, and stereotype. Two additional goals, both occupying a less central position in this cycle of the research agenda, are mentioned. The IAT, it is to become a widely useable technique to study implicit social cognition more generally, will require psychometric attention. Therefore, a third goal of the proposed research is to secure the computerized procedure and scaling properties of the measure. Finally, it is a goal of the proposed research to attempt basic research with a clear eye toward the application of implicit attitudes, stereotypes, and self-esteem findings. The scope of the present grant restricts our ability to develop these ideas here. Hence, in the present proposal we will only indicate the path to such opportunities and preparation for future independent development.

This project will fund the travel of the PI and two graduate students to present their research and participate in the European Social Cognition Network (ESCON) conference to be held in Brno, Czech Republic in September 2007. The ESCON meeting provides an international forum for the exchange of the most recent scientific developments in the field of social cognition. This conference is innovative not only in its cross-national representation, but also in its provision of a platform for young researchers -- including the two students whose travel will be supported on this grant -- to present their research ideas and to receive extensive feedback from a variety of experts in the field. The PI, as one of the invited plenary speakers, will present her research on the origins of social cognition to an international audience and will have the opportunity to discuss her work more broadly and informally with attendees. One of the broader goals of this meeting is to foster international research collaborations and to develop additional opportunities for young researchers to become integrated into the international scientific community. Participation in this conference will both enhance the research careers of the PI and graduate students, and will establish intellectual and social cross-national connections that will have a positive impact on the broader field of social cognition.

Psychological scientists currently know little about how it is that one person infers the mental states of others even though successful social interaction requires that we be capable of inferring the content of another person's mind, an ability known as "mentalizing." This research involves the continued integration of theoretical questions from social psychology with the methods and tools of cognitive neuroscience. The first objective is to apply the methods of cognitive neuroscience to examine how perceivers mentalize about other people who differ from them along the "Big Three" dimensions of age, race, and sex. Researchers have suggested that two different routes to mentalizing may exist. First, perceivers may sometimes be able to "simulate" another person's mental states by imagining themselves in the same situation and assuming that others would share the same thoughts and feelings that they experience themselves. However, this simulationist strategy is only appropriate when the other person is sufficiently similar to self that he or she would indeed experience the same mental states as oneself. If the other person is dissimilar to the self (e.g., when someone has very different opinions or a very different cultural background0 perceivers frequently rely upon "precompiled" social knowledge in the form of stereotypes. Recently, researchers have used brain imaging techniques (e.g., fMRI) to suggest that a particular brain region : the medial prefrontal cortex (MPFC) : may show a tradeoff between mentalizing based on simulation and mentalizing based on stereotyping. The current project investigates the nature of this "division of labor" in the MPFC. In addition, this research examines the changes produced by several strategies for reducing stereotyping, as well as the dynamics associated with learning about the similarity (or dissimilarity) of another person to oneself. By illuminating the brain basis of social cognition, it represents a first step towards understanding the social dysfunctions evidenced by patients with disorders such as autism, conduct disorder, or sociopathy. The research also has the potential to contribute to societal attempts at reducing intergroup bias by suggesting new forms of education and public policy based on an enhanced understanding of the neural basis of stereotyping.