Academic journal article Memory & Cognition

How Does Repetition Affect Memory? Evidence from Judgments of Recency

Academic journal article Memory & Cognition

How Does Repetition Affect Memory? Evidence from Judgments of Recency

Article excerpt

Four experiments were done to investigate the effects of repetition on judgment of recency (JOR). Experiment 1 showed that repetition can make an item seem either more recent or less recent than a nonrepeated item, depending on presentation spacing. Experiments 2-4 showed that subjects are able to judge the recency of a repeated item's first presentation or of its second presentation with a high degree of independence, especially if they report that the item occurred twice. The data are more consistent with an independent-trace explanation of JOR and repetition than with a cumulative-strength account, but neither hypothesis explains how repetition can make an item seem less recent. It is proposed that the findings as a whole can be better explained by a hypothesis based on recursive reminding.

After more than a century of research on episodic memory, there is still little agreement on mechanisms that underlie the effects of repetition. The most commonly considered alternatives are the cumulative-strength hypothesis and the multiple-trace hypothesis. The first of these assumes that when an experience is repeated, the memory representation that was formed during the first such experience is strengthened. The second assumes that each experienced event leaves behind its own separate memory trace, even if the event is a repetition.

The cumulative-strength versus multiple-trace distinction goes back at least to the 19th century (Ward, 1893), but both hypotheses are prominent among modern cognitive theories of memory. Examples of cumulative-strength theories can be found in McClelland and Chappell (1998), Murdock (1982), Murdock, Smith, and Bai (2001), and Wickelgren (1972). Examples of multiple-trace theories include Bower (1967), Hintzman (1986, 1988), Lansdale and Baguley (2008), and Logan (1988). Some theorists have vacillated between the two positions. In introducing the SAM model, Gillund and Shiffrin (1984) assumed that massed repetitions strengthen a single trace, whereas spaced repetitions may produce multiple traces; Shiffrin and Steyvers (1997) implemented the cumulative-strength assumption in their REM.1 model but used the multipletrace assumption in their REM.3 model; and Malmberg and Shiffrin (2005) discussed SAM and REM as though both were cumulative-strength models.

Repetition generally improves performance in standard recall and recognition-memory experiments, so for the purpose of explaining performances in these tasks, the cumulative-strength versus multiple-trace distinction may not much matter. Researchers investigating this issue have therefore turned to tasks that require other kinds of memory judgments. Judgments of recency (JOR) have played an especially important role in this work. There are two basic JOR procedures. In the forced-choice JOR or recencydiscrimination task, the experimental subjects choose the member of a test set (usually two items) that seems more recent. In the numerical JOR or absolute-judgment task, the subjects judge the number of items that intervened since a single test item was last presented. Data from both JOR tasks suggest that apparent recency approximately follows a logarithmic function of time or actual recency (e.g., Hinrichs, 1970; Yntema & Trask, 1963).

The conclusions that researchers have reached regarding repetition's effects have depended, in part, on which type of JOR test was used in collecting the data. In an early study, Morton (1968) required subjects to discriminate the recencies of two test digits, A and B, from a short preceding list. Taking B as the correct answer, the study conditions of interest were AB and AAB. (In the second condition, A occurred two times followed by a single occurrence of B.) Recency discrimination was less accurate in the AAB condition than in the AB condition-that is, the subjects had a greater tendency to choose the incorrect item if it had been repeated. Morton concluded from this result that remembered recency is based on strength and that repetition increases the strength of the item's memory (see also Murdock et al. …

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.