A retrieval model for both recognition and recall.

Abstract

The Search of Associative Memory (SAM) model for recall (Raaijmakers & Shiffrin, 1981b) is extended by assuming that a familiarity process is used for recognition. The recall model posits cue-dependent probabilistic sampling and recovery from an associative network. Our recognition model is closely related to the recall model because the total episodic activation due to the context and item cues is used in recall as a basis for sampling and in recognition to make a decision. The model, formalized in a computer simulation program, correctly predicts a number of findings in the literature as well as the results from a new experiment on the wordfrequency effect."A critical problem of long standing in psychological study of memory is concerned with the relation between recall and recognition. In what sense are they the same, and in what sense are they different?" (Tulving & Watkins, 1973, p. 739).This article is a preliminary attempt to formulate a theory that describes in detail the relationship between recall and recognition. The relationship is realized in a computer simulation model, and the model's predictions are checked against existing data as well as data generated in our laboratory. The model for recognition is related mathematically and logically to the Search of Associative Memory (SAM) theory of memory retrieval that has

A dynamic approach to recognition memory.

“…Separate-trace or exemplar theories (e.g., Gillund & Shiffrin, 1984;Hintz-man, 1988;Shiffrin & Steyvers, 1997;McClelland & Chappell, 1998) postulate that a distinct memory element is created each time an event is experienced. These memory elements-we will adopt the term "memory trace"-contain information about both the content of the event (e.g., the identity of a word on a study list) and the context in which the event occurred (where "context" can be defined at various levels, e.g., the time and location of the event, encoding task, and/or other nearby items or events; see Klein, Shiffrin, & Criss, 2007).…”

“…This comparison can take many forms, including a dot product (Hintzman, 1988) or likelihood ratio (Shiffrin & Steyvers, 1997;McClelland & Chappell, 1998); in any case, the comparison results in a single value reflecting the degree to which each memory trace is similar to the memory probe. This comparison is often assumed to take place in parallel across all memory traces (Gillund & Shiffrin, 1984;Hintzman, 1988;Shiffrin & Steyvers, 1997;McClelland & Chappell, 1998), but may take place over the course of a serial search among the various memory traces (Sternberg, 1966) or via a combination of parallel matching and serial search (Atkinson & Juola, 1974). Finally, these various match values are treated as evidence for whether the test item was or was not experienced in the target study context:…”

“…Most of the separate-trace models reviewed above assume that the chief source of noise is other items experienced within the target study context. That is, context cues alone are generally sufficient to "narrow down" the set of memory traces that are relevant for the decision, such that the memory probe is compared only with traces stored in the target context (Gillund & Shiffrin, 1984;Hintzman, 1988;Shiffrin & Steyvers, 1997;McClelland & Chappell, 1998) and noise results from the fact that different item can still partially match (a novel item might partially match one or more study items, while a studied item might also match some of the other studied item traces). The same is true for the TODAM (Murdock, 1982) and CHARM (Metcalfe Eich, 1982) models, which employ a composite-trace representation.…”

“…If the word frequency effect depends on first determining the frequency class to which the word belongs such that an appropriate criterion can be set (e.g., Gillund & Shiffrin, 1984;Glanzer & Adams, 1990;Hintzman, 1994), then the canonical mirror effect should only obtain when enough time is allowed for criterion setting. Similarly, if the LF hit rate advantage depends on recalling a LF target trace (Reder et al, 2000), this should take additional time to manifest.…”

“…Evidence from various recognition paradigms has supported the development of a variety of theories of memory (e.g., Atkinson & Juola, 1974;Metcalfe Eich, 1982;Murdock, 1982;Gillund & Shiffrin, 1984;Hintzman, 1988;Shiffrin & Steyvers, 1997;McClelland & Chappell, 1998;Dennis & Humphreys, 2001). These theories are concerned primarily with recognition accuracy and confidence ratings and can be viewed as extensions of a signal detection approach to recognition (Green & Swets, 1966).…”