Showing papers on "Fuzzy-trace theory published in 2007"

Converging evidence supports fuzzy-trace theory's nested sets hypothesis, but not the frequency hypothesis

Abstract: Evidence favors the nested sets hypothesis, introduced by fuzzy-trace theory (FTT) in the 1990s to explain “class-inclusion” effects and extended to many tasks, including conjunction fallacy, syllogistic reasoning, and base-rate effects (e.g., Brainerd & Reyna 1990; Reyna 1991; 2004; Reyna & Adam 2003; Reyna & Brainerd 1995). Crucial differences in mechanisms distinguish the FTT and Barbey & Sloman (B&S) accounts, but both contrast with frequency predictions (see Reyna & Brainerd, in press).

Effects of Inter-associate Connectivity on the Persistence of False Recall

Abstract: Effects of Inter-associate Connectivity on the Persistence of False Recall Winston D. Goh (psygohw@nus.edu.sg) Bee Hong Khoo (seventh@singnet.com.sg) Department of Psychology, National University of Singapore 11 Law Link, Singapore 117570, Singapore Abstract False recall of critical non-presented words in lists of associated words was examined with immediate testing and after a one-week delay. The word lists were manipulated such that the density of the inter-associate connectivity – the number pre-existing connections between the associates of the critical non-presented target – was either high or low. When recall occurred immediately after the study phase, high connectivity lists facilitated veridical recall compared to low connectivity lists, but connectivity had no effect on false recall. When recall was delayed for a week, low connectivity lists induced more false recall than high connectivity lists, but connectivity had no effect on veridical recall. The overall pattern of results is broadly consistent with the predictions of the Processing Implicit and Explicit Representations model of associative memory and not with the predictions of Fuzzy Trace Theory. Keywords: False recall; connectivity effects; retention interval; associative memory; long-term memory. Introduction More than a decade ago, Roediger and McDermott (1995) adapted Deese’s (1959) original list learning procedure and introduced what is now known as the Deese-Roediger- McDermott (DRM) paradigm for producing false memory in the laboratory. In this paradigm, participants are presented a study list of words (e.g. bed, rest, awake, tired, dream) which are the strongest associates of a critical target (e.g. sleep) that was not presented. The extent of false memory is indexed by the probability of remembering the critical non-presented target in a subsequent free recall or recognition test. Research on false memory using this paradigm or variants of it has been extensive. However, only a handful of studies have explicitly examined the persistence of false memory over time (e.g. McDermott, 1996; Thapar & McDermott, 2001; Toglia, Neuschatz, & Goodwin, 1999). The common finding among these studies is that false memory tends to persist over time, although the specific trends differ. McDermott (1996) reported that false recall exceeded veridical recall after a two-day retention interval. Toglia et al. (1999) reported that the level of false memory was not affected by retention intervals that varied between 0, 1, and 3 weeks, athough veridical recall declined over time. Thapar and McDermott (2001) reported that both veridical and false memory declined over retention intervals of 0, 2, and 7 days, but the rate of decline for veridical memory was more pronounced than that of false memory. Across these studies, veridical memory declined over time, as expected. Despite the inconsistent trends for false memory rates, taken together, the general pattern of results strongly suggests that the false memory effect using the DRM lists is robust and persistent over time. One theoretical account that has been used to explain the persistence of false memory is Fuzzy Trace Theory (Reyna & Brainerd, 1995; Brainerd & Reyna, 2002). In this theory, two types of independent memory representations are extracted in parallel – verbatim and gist traces. Verbatim traces are representations of the surface features of an experienced word and the concurrent contextual cues. Gist traces include overall patterns of the experience during list presentation, including meaning, relations, and patterns that are retrieved during the encoding of the words. When DRM lists are presented, an item-specific verbatim trace is created for each word. At the same time, the overall theme of each list is extracted and contributes to the gist traces. Memory for a word is dependent on both verbatim and gist traces. However, when one of the other trace is stronger, retrieval can be predominantly based on the stronger trace. Veridical memory is based on both types of representations, but false memory of critical non-presented targets can only arise from the gist traces since there is no verbatim trace. Gist and verbatim traces are assumed to have different rates of forgetting, with gist traces being more stable while verbatim traces fade more quickly. The loss of verbatim traces over time leads to the decline of veridical memory. As false memory of critical targets is dependent on the more stable gist traces, false memory is predicted to persist. An alternative account of false memory has been suggested that is based primarily on the idea of activations among associative networks, but which also relies on having two types of memory representations. The Processing Implicit and Explicit Representation (PIER 2) model of associative memory (Nelson, McKinney, Gee, & Janczura, 1998; Nelson, Schreiber, & McEvoy, 1992) assumes that the encoding of a word generates two independent memory representations – explicit and implicit representations. An explicit representation contains contextual information of the presentation of the word and its relation to other presented items, and arises through conscious processing. An implicit representation is created when the word’s lexical representation and its network of inter-related associates in long-term working memory are automatically and unconsciously activated. The spreading activation metaphor is used to describe the encoding of words by the implicit processing component. The sum activation of the encoded word depends on the number and strength of pre- existing connections between the word and its associates, as