Many researchers have hypothesized that recognition judgments are a result of two separate processes: a rapid familiarity-based process and a slower recall-like process. That is, recognition judgments that must be made very quickly are generally thought to be based on how familiar the test item seems, whereas judgments that are made more slowly may be partially or entirely based on a search of memory for the test item itself. The slower recall process has usually been assumed to function as a recall-to-reject process in which mismatching information that is retrieved from memory is used to reject a test probe. Recently, however, Heit and Jones (1996) tested the recall-to-reject account of item recognition and found no support for such a mechanism. In contrast, they found evidence consistent with a recall-to-accept process in which matching information is retrieved from memory to facilitate acceptance of a studied item. The current research extends Heit and Jones's work into the domain of associative recognition, in which two items or events are recognized as having occurred together in the past. Such associative recognition judgments have been thought to be based on a slow recall-to-reject process in addition to the fast familiarity-based process, for intuitively simple reasons. Recalling, for example, that John is married to Susan allows one to reject the idea that John is married to Mary. The specific nature of the recall process in associative recognition has important implications for the development of mathematical models of memory. Currently, many of those models assume that recognition judgments of both items and associations are based exclusively on the overall familiarity of the test item(s). The more familiar something seems, the more likely it is to be judged as `old;` as something that was experienced before. However, several lines of research have demonstrated that such a simple process cannot account for all recognition judgments. Clearly, then, the models need to be revised. The manner in which the models should be updated is not entirely obvious though, in part because we do not have adequate information about the operation of the recall component of recognition. The data that will be collected in the current research will inform memory modelers about the specific nature of that recall process. Armed with such information, our models can be revised in appropriate ways, with a minimum of false-starts, and our understanding of memory in general will be enhanced. This research meets the objectives of the POWRE program in two ways. First, it explores a very new aspect of memory. Although previous researchers have hypothesized that two processes might be involved in recognition, until recently no one has evaluated the specific nature of the recall-like component. Second, the investigator is at a critical point in her career regarding tenure consideration. The POWRE grant will allow for increased productivity, thus enhancing the investigator's potential for advancement and leadership in her field.
