Abstract
Previous work has shown that when items in a perceptual identification task are presented too quickly to identify, participants can still discriminate between studied and unstudied items. Such recognition without perceptual identification (RWPI) has been shown to occur in a variety of situations, including the false recognition of semantic associates of studied items. The present study investigated the utility of the RWPI paradigm for isolating instances of recognition that are familiarity based from those that are recollection based. Toward this end, the magnitude of the RWPI effect was compared in item versus associative recognition and in short versus long lists. The RWPI effect was larger in item than in associative recognition, and larger with short than with long study lists. These results are interpreted within the context of a dual-process approach to recognition and support the notion that RWPI taps familiarity-based recognition.
Acknowledgments
This research was conducted while Anne M. Cleary was supported by the National Institute of Child Health and Human Development Mental Retardation Research Training Grant 5-T32-HDO7176-16.
Notes
Most of the studies suggesting that intact pairs are not greater in familiarity than rearranged pairs have presented the words within each test pair simultaneously (e.g., CitationCleary, Curran, & Greene, 2001). However, the nature of the present perceptual identification task seemed to require that one word be presented at a time at test. First, it is unclear how simultaneously flashing two words for perceptual identification would affect the RWPI effect. For instance, it is possible that a word must be centrally fixated in order for RWPI for a word to occur, and it is unclear whether RWPI would occur for words flashed in one's periphery (as might occur if one word is on the left, and one is on the right, or one is on the top, and one is on the bottom), particularly if the peripheral word goes unattended (due to focus on the centrally fixated word). Second, flashing the two test words simultaneously on each test trial would lead to a rather complicated set of data analyses: Instances in which both words were identified would comprise one category; instances in which only one word was identified would comprise another, and instances in which neither word was identified would comprise another. Therefore, to simplify matters, we chose the present method of first flashing one word for identification and then presenting the second word for the associative recognition judgement.