Evan Heit - US grants

This project aims to improve our understanding of the basic processes of human learning and memory by developing instance-based models of memory. These models have been used to explain categorization tasks by assuming that our mental representations of categories consist of sets of category members or exemplars. Instance models are extended here to explain more complex phenomena from the areas of learning, memory, and judgment. These phenomena include reasoning from chains of examples, conceptual combination, learning relations between multiple variables, and judgment of event frequency. In experimental tests of these extended models, subjects will study examples, such as personality trait descriptions for a group of people, then make probability or classification judgments about the traits. The long-term objective of this research is to develop instance-based models into a framework that helps psychologists understand a wide range of cognitive abilities.

Are there two kinds of reasoning? Informally, people often make a distinction between reasoning by intuition versus reasoning by logical deliberation. Intuition generates a "quick and dirty" answer, whereas logical deliberation takes time and effort but yields a more justifiable answer. Despite the appeal of this distinction, the scientific evidence for it is lacking. Most psychological theories of reasoning either focus on only one kind of reasoning, or assume that there are two kinds but do not spell out the details of how they are applied in different situations. However, in other areas of psychology such as memory research, the hypothesis of complementary slow and fast systems has been pursued in depth with considerable success.

With support of the National Science Foundation, Dr. Heit and Dr. Rotello will study reasoning using techniques that have been adapted from memory research. For example, participants in experiments will solve reasoning problems with different time deadlines: The intuitive system is expected to dominate quicker judgments and the deliberative system is expected to dominate slower judgments. The program of empirical research is complemented by the development of mathematical models of reasoning processes. These models will help in designing reasoning problems that can be used to test the underlying theories, as well as assess reasoning skills. For instance, the problems will be administered to undergraduate students at the beginning and end of a course on critical thinking. The models will be used to identify which aspects of reasoning skill are improved by course instruction. This project will help to establish research partnerships between the psychology departments at the University of Massachusetts and the newest research university in the US, the University of California, Merced, which has a high proportion of first generation college students and other under-represented groups.

The interdisciplinary study of mind, brain, and thought --- Cognitive Science --- is about fifty years old. The field is thriving, producing insight into the mechanisms of thought by integrating observation, experimentation, computational modeling, and theory from a variety of traditional disciplines. But the place of Cognitive Science in education and the workforce is unclear, as students and professionals are driven towards ever more specialization and ever more grounding in technical disciplines. What are the prospects for Cognitive Science on the eve of its fiftieth birthday? What are the current, dominant theories and methods? How can Cognitive Science inform the design of human technologies and tools? How should education and training in Cognitive Science evolve to meet the needs of the 21st century workforce?

The National Science Foundation will sponsor a conference to take stock of the current state of Cognitive Science, and to look to its future in the academic and industry settings. Cognitive scientists from universities, government agencies, and a variety of companies and organizations from across the US will participate. Participants will be chosen to cover the gamut of cognitive science expertise, including experimentalists, modelers, computer scientists, language theorists, ethnographers, and educators. The conference will be held at UC Merced, a new research university with plans to develop a world-class Cognitive Science department.