Reinforcement-Based Models of Delayed Matching-to-Sample Performance

点击翻译按钮获取中文摘要

Dr. Wixted's research is designed to enhance our understanding of the principles that govern animal memory. If successful, the results may help to solve the far more complex puzzle of how memory operates, and sometimes goes awry, in humans. The basic procedure used throughout these experiments involves presenting a series of trials, during each of which a color (e.g., red) is shown to a pigeon in a center window for a few seconds, the color is removed for a period of time, and then the pigeon is given a choice between two colors (red vs. green) in two side windows. A response to the matching color (in this case, red) is rewarded with food while a response to the nonmatching color simply ends the trial. Dr. Wixted will carry out three sets of experiments that address different, but related, issues. The first set is concerned with the question of how short-term and long-term memory compete to determine the pigeon's choice on a given trial. More specifically, the choice between red and green may involve a competition between short-term memory for the previous color (red) and long-term memory for the choice color that generally provides the greater number of rewards. The second set of experiments is concerned with identifying the psychological variables that cause a stimulus to become more (or less) memorable. The principle under investigation holds that a stimulus associated with a short delay to a biologically significant event (e.g., food) is more memorable than one associated with a long delay. Finally, the third set of experiments investigates the variables that influence memory for nonoccurrence. That is, under what conditions will an animal remember that no stimulus appeared in the center window on a particular trial? Together, the results of these experiments will help to identify the principles that underlie animal and, ultimately perhaps, human memory.