Facial appearance strongly influences first impressions, which show remarkable consensus and yield significant social consequences. The long-range objective is to explain consensual impressions among normal individuals and their absence in individuals with disorders such as autism. It is hypothesized that qualities that are accurately revealed by facial features that mark babies, unfitness, or identity are overgeneralized to people whose facial structure resembles that of babies, a particular level of fitness, or a particular identity. Four specific research aims are: 1)connectionist modeling tests the facial identity overgeneralization hypothesis (FIOH) that the tendency for responses to strangers to vary with their facial resemblance to known individuals contributes to social categorization effects; 2)functional Magnetic Resonance Imaging (fMRI) investigates brain activation patterns in response to faces that are predicted from each of the overgeneralization hypotheses; 3)generalized mere exposure paradigms investigate whether familiarization with the facial prototype of a category changes impressions of new category members, as predicted by the FIOH; 4)virtual reality methods test face overgeneralization hypotheses through behavioral responses to faces in various social contexts. Four training aims are: 1) increased knowledge of the conceptual foundations of computer modeling of neural systems and the computer software used to implement neural network simulations; 2) increased knowledge of brain structure and function and fMRI techniques; 3) learning to use virtual environments to assess the perceived behavioral affordances of different facial qualities; 4) integrating various methods to provide more sophisticated tests of the face overgeneralization hypotheses. The research and training benefit from excellent laboratory facilities, state of the art equipment, and interactions with outstanding scientists at Brandeis University and the MGH-NMR Center. Integrating social psychology, computer science, and neuroscience promises to advance the understanding of social perception. The focus on facial information offers a novel theoretical perspective, demonstrating that intrinsic properties of faces contribute to social biases that have been largely viewed as social constructions. It also provides a foundation for research that may elucidate neural bases of social deficits, as in autism, and develop intervention techniques to modify reactions to faces.
