Academic journal article Memory & Cognition

Looking for Graded Recollection: Manipulating the Number of Details to Be Recollected Does Not Affect Recollection Variance

Academic journal article Memory & Cognition

Looking for Graded Recollection: Manipulating the Number of Details to Be Recollected Does Not Affect Recollection Variance

Article excerpt

Abstract Recollection has been the subject of much debate, with some models maintaining that it is subject to a threshold, some maintaining that it is a continuous process, and some maintaining that both are true. Threshold models maintain that recollection can fail (i.e., fall below threshold), whereas signal detection models treat recollection as a continuous process. Recent research has revealed that some manipulations can influence this behavior, but the general reasons why threshold patterns emerge in some conditions and graded patterns emerge in others are still unclear. One potential explanation is the number of retrieved details; recollection of stimuli with few details may succeed or fail, whereas recollection of stimuli with many details may be graded. If this is true, manipulating the amount of detail should produce threshold patterns for "few details" conditions, but more graded patterns for "many details" conditions. This hypothesis was tested in six experiments examining source memory, and the number-ofdetails manipulation consistently failed to affect the nature of recollection. Overall, the results suggest that the amount of information available to be retrieved does not, by itself, explain recollection variability.

Keywords Recognition . Recollection . Dual-process signal detection . Signal detection theory . Source memory . Threshold

Recognition memory is largely agreed to rely on two processes, recollection and familiarity (Atkinson & Juola, 1974; Jacoby, 1991; Mandler, 1980; Yonelinas, 1994). Over the past several years, debates about the operation of recognition memory have shifted from arguments about the number of processes to debates about how the processes operate (e.g., C. M. Parks & Yonelinas, 2007, 2009; Rotello, Macmillan, & Reeder, 2004; Rotello, Macmillan, Reeder, & Wong, 2005; Slotnick & Dodson, 2005; Wixted, 2007) and the neural substrates that are most critical to each process (e.g., Diana, Yonelinas, & Ranganath, 2007; Squire, 2004; Yonelinas et al., 2002). Traditionally, dual-process theory defines recollection as the recall of qualitative episodic detail about a prior event, and familiarity as a quantitative strength signal indicative of the "oldness" of a stimulus. Many researchers agree that signal detection theory (SDT; e.g., Macmillan & Creelman, 2005) provides a good account of familiarity at both a quantitative modeling level and a conceptual level. However, there is still widespread disagreement about recollection; although most researchers are content with the idea that recollection reflects retrieval of qualitative information about prior events, debates about how to describe that form of memory in process models are ongoing. Currently, one major point of contention is whether recollection is a continuous or a threshold process.

In line with traditional definitions of recollection (e.g., Atkinson & Juola, 1974; Jacoby, 1991; Mandler, 1980), the dual-process signal detection (DPSD) model treats recollection as a threshold process and familiarity as a signal detection process (Yonelinas, 1994; Yonelinas & Parks, 2007). One interpretation of the unequal-variance signal detection (UVSD) model (Wixted, 2007) also includes recollection and familiarity at a conceptual level, but assumes that they operate in the same continuous fashion; thus, both recollection and familiarity are considered to be signal detection processes subject to criteria setting, but not to a threshold. The fundamental difference between the threshold and criterion ideas is that variations in the signal below a threshold are uninterpretable and nondiagnostic-they are, essentially, useless, and thus represent a failure of memory. In contrast, variations below a criterion could still offer useful fill information to the participant. This difference was demonstrated by C. M. Parks and Yonelinas (2009) in experiments using the second-choice procedure, in which participants chose "back-up" responses, many of which would fall below threshold or below criterion. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.