Academic journal article Canadian Journal of Experimental Psychology

Dissociations of Processes in Recognition Memory: Effects of Interference and of Response Speed

Academic journal article Canadian Journal of Experimental Psychology

Dissociations of Processes in Recognition Memory: Effects of Interference and of Response Speed

Article excerpt

Abstract Effects on two bases for recognition-memory judgements were examined using a process dissociation procedure (Jacoby, 1991). In three experiments it was found that increasing the length of a study list interfered with conscious recollection but left familiarity in place. Furthermore, an examination of reaction time distributions as well as results from a response-signal procedure showed that familiarity was faster as a basis for recognition judgements than was conscious recollection. However, both bases contributed to performance on the fastest as well as the slowest responses, suggesting that the two processes were acting in parallel.

For a test of recognition memory, subjects must decide whether a test item was presented in a previously studied list. At least in principle, subjects could base their recognition judgements solely on the familiarity of the test items because, on average, an item that was presented in the study list would be more familiar than one that was not. A rationale of this sort underlies single-factor theories such as signal detection theory (see Swets, 1964; Wickelgren, 1972). However, subjects may not be limited to assessments of item familiarity when making recognition-memory judgements. If some aspect of the study event could be consciously recollected (e.g., 'I remember seeing that word... it was the first word in the list') this could serve as a second basis for responding.

Several researchers have proposed a dual-process view of recognition memory along with criteria that can supposedly be used to distinguish between the two bases for responding. Familiarity is assumed to be a fast basis for responding. Familiarity is assumed to be a fast basis for responding (Atkinson & Juola, 1974; Jacoby, 1991; Mandler, 1980) that relies on perceptual characteristics (Jacoby & Dallas, 1981; Mandler, 1980) or item-specific information (Humphreys & Bain, 1983), and reflects an automatic or unconscious use of memory (Jacoby, 1991) that is largely spared by amnesia (e.g., Piercy & Huppert, 1972; Verfaellie & Treadwell, 1993). In contrast, the use of recollection is described as a slow, search-like process that relies on conceptual processing or associative information and requires attention. Furthermore, recollection is said to be absent or reduced in amnesic patients.

In this paper we examine the two bases for recognition memory by focussing on differential effects of interference, and differences in response speed. Although the examination of interference effects has a long history in recognition experiments, it has received little attention in the context of dual-process theories. Thus we know that manipulations such as increasing the length of the study list (e.g., Strong, 1912) or increasing the delay between study and test (e.g., Strong, 1913) interferes with recognition memory performance, but we do not known whether just one or both of the two bases for recognition are influenced.

To examine interference effects, we varied the length of study lists with the expectation that increasing list length would increase interference. The list length effect was first reported by Strong (1912) who found that as he increased the number of advertisements a subject studied, the probability of later recognizing a particular advertisement decreased. Since then the list length effect has been demonstrated numerous times in recognition (e.g., Atkinson & Shiffrin, 1971; Gillund & Shiffrin, 1984; Ratcliff & Murdock, 1976), free recall (Murdock, 1960; Roberts, 1972) and word-fragment completion (Sloman, Hayman, Ohta, Law & Tulving, 1988).

A number of observations led us to believe that recollection might be more susceptible to interference than familiarity. First, the magnitude of the list-length effect is greater in free recall than in recognition. Doubling list length produces a dramatic drop in free recall (Roberts, 1972) but only a small decrease in recognition (Ratcliff & Murdock, 1976). …

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.