Academic journal article Memory & Cognition

Investigating the Encoding-Retrieval Match in Recognition Memory: Effects of Experimental Design, Specificity, and Retention Interval

Academic journal article Memory & Cognition

Investigating the Encoding-Retrieval Match in Recognition Memory: Effects of Experimental Design, Specificity, and Retention Interval

Article excerpt

Five experiments investigated the encoding-retrieval match in recognition memory by manipulating read and generate conditions at study and at test. Experiments 1A and 1B confirmed previous findings that reinstating encoding operations at test enhances recognition accuracy in a within-groups design but reduces recognition accuracy in a between-groups design. Experiment 2A showed that generating from anagrams at study and at test enhanced recognition accuracy even when study and test items were generated from different anagrams. Experiment 2B showed that switching from one generation task at study (e.g., anagram solution) to a different generation task at test (e.g., fragment completion) eliminated this recognition advantage. Experiment 3 showed that the recognition advantage found in Experiment 1A is reliably present up to 1 week after study. The findings are consistent with theories of memory that emphasize the importance of the match between encoding and retrieval operations.

The view that memory accuracy is determined by the match between encoding and retrieval processes is one of the cornerstones of modern memory research. Theories such as the transfer-appropriate processing framework (Morris, Bransford, & Franks, 1977), the encoding specificity principle (Tulving & Thompson, 1973), and the procedural approach to memory (Kolers, 1973, 1975) all emphasize the importance of the encoding-retrieval match (for reviews, see Nairne, 2002; Roediger, Gallo, & Geraci, 2002; Roediger & Guynn, 1996). More recently, Kent and Lamberts (2008) suggested that the encoding- retrieval match relies on the mental simulation of encoding processes (see also Barsalou, 2008). According to these theories, records of the cognitive operations engaged at encoding are integrated with the information acquired via those operations. Reinstating the same operations at test cues the retrieval of the acquired information, leading to an increase in recognition accuracy. The aim of the present study was to identify some of the boundary conditions of this effect. Specifically, we investigated the effect of experimental design (within-groups versus betweengroups manipulations), the specificity with which encoding and retrieval processes overlap, and the duration of the recognition advantage conferred by reinstating encoding operations at test.

The importance of the encoding-retrieval match has been demonstrated in a number of studies in which the orienting task carried out at study was performed again at test. For example, Glisky and Rabinowitz (1985, Experiment 1) presented participants with five-letter words that were either read intact or generated from four-letter fragments. They found that the generation effect (greater recognition accuracy for words that were generated rather than read; Slamecka & Graf, 1978) was enhanced when participants also had to generate the test items prior to the recognition decision. This finding was replicated by Dewhurst and Brandt (2007), who found that reinstatement of the generation task at test selectively enhanced the conscious recollection of studied items, as indicated by an increase in remember responses, but not one in know responses (Gardiner, 1988; Tulving, 1985).

A further demonstration of the encoding-retrieval match was reported by Engelkamp, Zimmer, Mohr, and Sellen (1994) using the enactment paradigm. Participants read a series of action phrases, such as "close the book," and on some trials were also instructed to perform the action. Consistent with the enactment effect, participants remembered enacted phrases better than they remembered phrases that were only read. Engelkamp et al. found that this effect was enhanced when participants performed the actions again prior to the recognition decision (see also Mulligan & Hornstein, 2003, who found an advantage when both participant-performed and experimenter-performed tasks were reinstated at test). Considered together, the findings above provide powerful support for the view that recognition accuracy is significantly enhanced when the operations engaged at study are reinstated at test. …

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.