Search

Attentional subsetting theory (Caplan, 2023) posits that only a small subset of item features are attended in episodic recognition tasks. This explained a pivotal finding for the development of recognition models: the near-null list-strength effect, where encoding strength influences recognition similarly in mixed-strength lists and pure-strength lists. Most research uses spaced repetition to manipulate encoding strength. However, the origin of the null list-strength effect was a more unusual manipulation of stimulus duration (1 s versus 2 s) — and reported an inverted list-strength effect. We present an attentional subsetting theory of duration that produces inversions — and explains why they are uncommon: Earlier-attended features dwell within a lower-dimensional feature subspace, which participants can sometimes disregard during test trials of pure-strong lists, giving strong-pure items an extra advantage. The model previously only solved for d ′. We extend it to generate realistic hit and false-alarm rates by deriving the criterion from attention to each probe. Supporting the theory, two pre-registered experimental manipulations of stimulus-duration reproduced robust inverted list-strength effects, suggesting this type of finding is unlikely due to sampling error. This account of stimulus-duration, explaining inverted, as well as upright and null, list-strength effects, could be incorporated in most models with vector representations • The theory: participants attend a small, idiosyncratic subset of an items' features. • Early versus later attended features may differ in dimensionality. • This explains how long study times can have more advantage in different lists. • Two experiments support this prediction. [ABSTRACT FROM AUTHOR]

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.jml.2010.03.001 Byline: Christopher R. Madan (a), Mackenzie G. Glaholt (b), Jeremy B. Caplan (a)(c) Keywords: Association-memory; Associative symmetry; Imageability; Word frequency; Episodic memory Abstract: Word properties like imageability and word frequency improve cued recall of verbal paired-associates. We asked whether these enhancements follow simply from prior effects on item-memory, or also strengthen associations between items. Participants studied word pairs varying in imageability or frequency: pairs were 'pure' (high-high, low-low) or 'mixed' (high-low, low-high) where 'high' and 'low' refer to imageability or frequency values and are probed with forward (A-?) and backward (?-B) cues. Probabilistic model fits to the data suggested that imageability primarily improved retrieval of associations, but frequency primarily improved recall of target items. All pair types exhibited a high correlation between forward and backward probe accuracy, a measure of holistic learning (), which extends the boundary conditions of holistic association-memory and challenges suggestion that holistic learning depends critically on imagery. In sum, item properties can boost association-memory beyond simply boosting target retrievability. Author Affiliation: (a) Department of Psychology, University of Alberta, Canada (b) Department of Psychology, University of Toronto at Mississauga, Canada (c) Centre for Neuroscience, University of Alberta, Canada Article History: Received 29 May 2009; Revised 16 February 2010

Association-memory is a major focus of verbal memory research. However, experimental paradigms have only occasionally tested memory for the order of the constituent items (AB versus BA). Published models of association-memory, implicitly, make clear assumptions about whether associations are learned without order (e.g., convolution-based models) or with unambiguous order (e.g., matrix models). Seeking empirical data to test these assumptions, participants studied lists of word-pairs, and were tested with cued recall, associative recognition and constituent-order recognition. Order-recognition was well above chance, challenging strict convolution-based models, but only moderately coupled with association-memory. Convolution models are thus insufficient, needing an additional mechanism to infer constituent order, in a manner that is moderately correlated with association-memory. Current matrix models provide order, but over-predict the coupling of order- and association-memory. In a simulation, when we allowed for order to be wrongly encoded for some proportion of pairs, order-recognition could be decoupled from cued recall. This led to the prediction that participants should persist with their incorrect order judgement between initial and final order-recognition, but this was not supported by the data. These findings demand that current models be amended, to provide order-memory, while explaining how order can be ambiguous even when the association, itself, is remembered. [ABSTRACT FROM AUTHOR]

Abstract: Emotionally arousing information is remembered better than neutral information. This enhancement effect has been shown for memory for items. In contrast, studies of association-memory have found both impairments and enhancements of association-memory by arousal. We aimed to resolve these conflicting results by using a cued-recall paradigm combined with a model-based data analysis method () that simultaneously obtains separate estimates of arousal effects on memory for associations and memory for items. Participants studied sequentially presented words in pairs that were pure (NEGATIVE–NEGATIVE or NEUTRAL–NEUTRAL) or mixed (NEGATIVE–NEUTRAL or NEUTRAL–NEGATIVE). Cued recall tests had NEUTRAL or NEGATIVE probes and NEUTRAL or NEGATIVE targets. We found impaired memory for associations involving negative words despite enhanced item-memory (more retrievable targets). A category-list control condition explained away the item-memory enhancement but could not explain the impairment of association-memory due to arousal. A second experiment with identical structure but using higher arousing taboo words revealed increased cued recall of taboo than neutral words. However, this was exclusively mediated by item-memory effects with neither enhancement nor impairment of association-memory. Thus, although cued recall was lower for pure negative pairs and higher for pure taboo pairs, our modeling approach determined a different locus of action for these memory impairing or increasing effects: Although item memory was increased by arousal, association-memory was impaired by moderately arousing, negative words and unaffected by taboo words. Our results suggest that previous results reporting an enhancement of association-memory due to arousal may have instead been solely driven by enhanced item-memory. [Copyright &y& Elsevier]

• Reward value influences memory. • We manipulate value while equating task-usefulness. • Value did not exert a large influence on free recall probability. • Value did influence recall order and lexical judgements. • High-value words were recalled earlier and responded faster. A common finding is that items associated with higher reward value are subsequently remembered better than items associated with lower value. A confounding factor is that when a higher value stimuli is presented, this typically signals to participants that it is now a particularly important time to engage in the task. When this was controlled, Madan, Fujiwara, Gerson, and Caplan (2012) still found a large value-bias of memory. Their value-learning procedure, however, explicitly pitted high- against low-value words. Our novel value-learning procedure trained words one at a time, avoiding direct competition between words, but with no difference in words signalling participants to engage in the task. Results converged on null effects of value on subsequent free recall accuracy. Re-analyses attributed Madan et al.'s value-bias to competition between choice items that were paired during learning. Value may not bias memory if it does not signal task importance or induce inter-item competition. [ABSTRACT FROM AUTHOR]