Search

The article reports four experiments with complex-span tasks in which encoding of memory items alternates with processing of distractors. The experiments test two assumptions of a computational model of complex span, SOB-CS: (1) distractor processing impairs memory because distractors are encoded into working memory, thereby interfering with memoranda; and (2) free time following distractors is used to remove them from working memory by unbinding their representations from list context. Experiment 1 shows that distractors are erroneously chosen for recall more often than not-presented stimuli, demonstrating that distractors are encoded into memory. Distractor intrusions declined with longer free time, as predicted by distractor removal. Experiment 2 shows these effects even when distractors precede the memory list, ruling out an account based on selective rehearsal of memoranda during free time. Experiments 3 and 4 test the notion that distractors decay over time. Both experiments show that, contrary to the notion of distractor decay, the chance of a distractor intruding at test does not decline with increasing time since encoding of that distractor. Experiment 4 provides additional evidence against the prediction from distractor decay that distractor intrusions decline over an unfilled retention interval. Taken together, the results support SOB-CS and rule out alternative explanations. Data and simulation code are available on Open Science Framework: osf.io/3ewh7. [ABSTRACT FROM AUTHOR]

We examine the explanatory roles that have been ascribed to various forms of rehearsal or refreshing in short-term memory (STM) and working memory paradigms, usually in conjunction with the assumption that memories decay over time if they are not rehearsed. Notwithstanding the popularity of the rehearsal notion, there have been few detailed examinations of its underlying mechanisms. We explicitly implemented rehearsal in a decay model and explored its role by simulation in several benchmark paradigms ranging from immediate serial recall to complex span and delayed recall. The results show that articulatory forms of rehearsal often fail to counteract temporal decay. Rapid attentional refreshing performs considerably better, but so far there is scant empirical evidence that people engage in refreshing during STM tasks. Combining articulatory rehearsal and refreshing as 2 independent maintenance processes running in parallel leads to worse performance than refreshing alone. We conclude that theoretical reliance on articulatory rehearsal as a causative agent in memory may be unwise and that explanatory appeals to rehearsal are insufficient unless buttressed by quantitative modeling. [ABSTRACT FROM AUTHOR]

The Hebb repetition effect refers to the finding that immediate serial recall is improved over trials for memory lists that are surreptitiously repeated across trials, relative to new lists. We show in four experiments that the Hebb repetition effect is also observed with a complex-span task, in which encoding or retrieval of list items alternates with an unrelated processing task. The interruption of encoding or retrieval by the processing task did not reduce the size of the Hebb effect, demonstrating that incidental long-term learning forms integrated representations of lists, excluding the interleaved processing events. Contrary to the assumption that complex-span performance relies more on long-term memory than standard immediate serial recall (simple span), the Hebb effect was not larger in complex-span than in simple-span performance. The Hebb effect in complex span was also not modulated by the opportunity for refreshing list items, questioning a role of refreshing for the acquisition of the long-term memory representations underlying the effect. [ABSTRACT FROM AUTHOR]

Highlights: [•] We introduce a method for isolating a process unique to working memory updating. [•] Removal of outdated information is an active working memory updating process. [•] Removal speed can be measured reliably. [•] Removal speed is independent of working memory capacity. [Copyright &y& Elsevier]

Recent research has found a positive relationship between people's working memory capacity (WMC) and their speed of category learning. To date, only classification-learning tasks have been considered, in which people learn to assign category labels to objects. It is unknown whether learning to make infer-ences about category features might also be related to WMC. We report data from a study in which 119 participants undertook classification learning and inference learning, and completed a series of WMC tasks. Working memory capacity was positively related to people's classification and inference learning performance. [ABSTRACT FROM AUTHOR]

The article tests the assumption that forgetting in working memory for verbal materials is caused by time-based decay, using the complex-span paradigm. Participants encoded 6 letters for serial recall; each letter was preceded and followed by a processing period comprising 4 trials of difficult visual search. Processing duration, during which memory could decay, was manipulated via search set size. This manipulation increased retention interval by up to 100% without having any effect on recall accuracy. This result held with and without articulatory suppression. Two experiments using a dual-task paradigm showed that the visual search process required central attention. Thus, even when memory maintenance by central attention and by articulatory rehearsal was prevented, a large delay had no effect on memory performance, contrary to the decay notion. Most previous experiments that manipulated the retention interval and the opportunity for maintenance processes in complex span have confounded these variables with time pressure during processing periods. Three further experiments identified time pressure as the variable that affected recall. We conclude that time-based decay does not contribute to the capacity limit of verbal working memory. [ABSTRACT FROM AUTHOR]

This article introduces a new computational model for the complex-span task, the most popular task for studying working memory. SOB-CS is a two-layer neural network that associates distributed item representations with distributed, overlapping position markers. Memory capacity limits are explained by interference from a superposition of associations. Concurrent processing interferes with memory through involuntary encoding of distractors. Free time in-between distractors is used to remove irrelevant representations, thereby reducing interference. The model accounts for benchmark findings in four areas: (1) effects of processing pace, processing difficulty, and number of processing steps; (2) effects of serial position and error patterns; (3) effects of different kinds of item-distractor similarity; and (4) correlations between span tasks. The model makes several new predictions in these areas, which were confirmed experimentally. [ABSTRACT FROM AUTHOR]

The concept of attention is central to theorizing in learning as well as in working memory. However, research to date has yet to establish how attention as construed in one domain maps onto the other. We investigate two manifestations of attention in category- and cue-learning to examine whether they might provide common ground between learning and working memory. Experiment 1 examined blocking and highlighting effects in an associative learning paradigm, which are widely thought to be attentionally mediated. No relationship between attentional performance indicators and working memory capacity (WMC) was observed, despite the fact that WMC was strongly associated with overall learning performance. Experiment 2 used a knowledge restructuring paradigm, which is known to require recoordination of partial category knowledge using representational attention. We found that the extent to which people successfully recoordinated their knowledge was related to WMC. The results illustrate a link between WMC and representational--but not dimensional--attention in category learning. [ABSTRACT FROM AUTHOR]

Working memory is crucial for many higher level cognitive functions, ranging from mental arithmetic to reasoning and problem solving. Likewise, the ability to learn and categorize novel concepts forms an indispensable part of human cognition. However, very little is known about the relationship between working memory and categorization. This article reports 2 studies that related people's working memory capacity (WMC) to their learning performance on multiple rule-based and information-integration perceptual categorization tasks. In both studies, structural equation model-ing revealed a strong relationship between WMC and category learning irrespective of the require-ment to integrate information across multiple perceptual dimensions. WMC was also uniformly related to people's ability to focus on the most task-appropriate strategy, regardless of whether or not that strategy involved information integration. Contrary to the predictions of the multiple systems view of categorization, working memory thus appears to underpin performance in both major classes of perceptual category-learning tasks. [ABSTRACT FROM AUTHOR]

There has been growing interest in the relationship between the capacity of a person's working memory and their ability to learn to categorize stimuli. While there is evidence that working memory capacity (WMC) is related to the speed of category learning, it is unknown whether WMC predicts which strategies people use when there are multiple possible solutions to a categorization problem. To explore the relationship between WMC, category learning, and categorization strategy use, 173 participants completed two categorization tasks and a battery of WMC tasks. WMC predicted the speed of category learning, but it did not predict which strategies participants chose to perform categorization. Thus, WMC does not predict which categorization strategies people use but it predicts how well they will use the strategy they select. [ABSTRACT FROM AUTHOR]

Working memory is crucial for many higher-level cognitive functions, ranging from mental arithmetic to reasoning and problem solving. Likewise, the ability to learn and categorize novel concepts forms an indispensable part of human cognition. However, very little is known about the relationship between working memory and categorization, and modeling in category learning has thus far been largely uninformed by knowledge about people's memory processes. This article reports a large study (N = 113) that related people's working memory capacity (WMC) to their category-learning performance using the 6 problem types of Shepard, Hovland, and Jenkins (1961). Structural equation modeling revealed a strong relationship between WMC and category learning, with a single latent variable accommodating performance on all 6 problems. A model of categorization (the Attention Learning COVEring map, ALCOVE; Kruschke, 1992) was fit to the individual data and a single latent variable was sufficient to capture the variation among associative learning parameters across all problems. The data and modeling suggest that working memory mediates category learning across a broad range of tasks. [ABSTRACT FROM AUTHOR]

Working memory is a core concept in cognition, predicting about 50% of the variance in IQ and reasoning tasks. A popular test of working memory is the complex span task, in which encoding of memoranda alternates with processing of distractors. A recent model of complex span performance, the Time-Based-Resource-Sharing (TBRS) model of Barrouillet and colleagues, has seemingly accounted for several crucial findings, in particular the intricate trade-off between deterioration and restoration of memory in the complex span task. According to the TBRS, memory traces decay during processing of the distractors, and they are restored by attentional refreshing during brief pauses in between processing steps. However, to date, the theory has been formulated only at a verbal level, which renders it difficult to test and to be certain of its intuited predictions. We present a computational instantiation of the TBRS and show that it can handle most of the findings on which the verbal model was based. We also show that there are potential challenges to the model that await future resolution. This instantiated model, TBRS*, is the first comprehensive computational model of performance in the complex span paradigm. The Matlab model code is available as a supplementary material of this article. [ABSTRACT FROM AUTHOR]

We investigated the effects of the duration and type of to-be-articulated distractors during encoding of a verbal list into short-term memory (STM). Distractors and to-be-remembered items alternated during list presentation, as in the complex-span task that underlies much of working-memory research. According to an interference model of STM, known as serial order in a box (SOB; Farrell & Lewandowsky, 2002), additional repeated articulations of the same word between list items should cause minimal further disruption of encoding into STM even though the retention interval for early list items is increased. SOB also predicts that the articulation of several different distractor items should lead to much enhanced disruption if the distractor interval is increased. Those predictions were qualitatively confirmed in 4 experiments that found that it is the type of distractors, not their total duration, that determines the success of encoding a list into STM. The results pose a challenge to temporal models of complex-span performance, such as the time-based resource sharing model (Barrouillet, Bernardin, & Camos, 2004). The results add to a growing body of evidence that memory for the short term is not exclusively governed by purely temporal processes. [ABSTRACT FROM AUTHOR]

We present a battery of four working memory tasks that are implemented using MATLAB and the free Psychophysics Toolbox. The package includes preprocessing scripts in R and SPSS to facilitate data analysis. The four tasks consist of a sentence-span task, an operation-span task, a spatial short-term memory test, and a memory updating task. These tasks were chosen in order to provide a heterogeneous set of measures of working memory capacity, thus reducing method variance and tapping into two content domains of working memory (verbal, including numerical, vs. spatial) and two of its functional aspects (storage in the context of processing and relational integration). The task battery was validated in three experiments conducted in two languages ( English and Chinese), involving more than 350 participants. In all cases, the tasks were found to load on a single latent variable. In a further experiment, the latent working memory variable was found to correlate highly but not perfectly with performance on Raven's matrices test of fluid intelligence. We suggest that the battery constitutes a versatile tool to assess working memory capacity with either English- or Chinese-speaking participants. The battery can be downloaded from www.cogsciwa.com ("Software" button). [ABSTRACT FROM AUTHOR]

Working memory updating (WMU) has been identified as a cognitive function of prime importance for everyday tasks and has also been found to be a significant predictor of higher mental abilities. Yet, little is known about the constituent processes of WMU. We suggest that operations required in a typical WMU task can be decomposed into 3 major component processes: retrieval, transformation, and substitution. We report a large-scale experiment that instantiated all possible combinations of those 3 component processes. Results show that the 3 components make independent contributions to updating performance. We additionally present structural equation models that link WMU task performance and working memory capacity (WMC) measures. These feature the methodological advancement of estimating interindividual covariation and experimental effects on mean updating measures simultaneously. The modeling results imply that WMC is a strong predictor of WMU skills in general, although some component processes-in particular, substitution skills-were independent of WMC. Hence, the reported predictive power of WMU measures may rely largely on common WM functions also measured in typical WMC tasks, although substitution skills may make an independent contribution to predicting higher mental abilities. [ABSTRACT FROM AUTHOR]

What drives forgetting in working memory? Recent evidence suggests that in a complex-span task in which an irrelevant processing task alternates with presentation of the memoranda, recall declines when the time taken to complete the processing task is extended while holding the time for rehearsal in between processing steps Constant (Portrat, Barrouillet, & Camos, 2008). This time-based forgetting was interpreted in support for the role of time-based decay in working memory. In this article, we argue the contrary position by (a) showing in an experiment that the processing task in Portrat et al's (2008) study gave rise to uncontrolled post-error processes that occupied the attentional bottleneck, thus preventing restorative rehearsal, and (b) showing that when those post-error processes are statistically controlled, there is no evidence for temporal decay in Porirat et al's study. We conclude that currently there exists no direct evidence for temporal decay in the complex-span paradigm. [ABSTRACT FROM AUTHOR]

Many models of short-term memory (STM) ascribe an important role to temporal decay and forgetting because of the passage of time alone. We argue against decay as the primary form of forgetting from STM, and suggest that new experimental methodologies and recent models provide new perspectives on the old issue of the causes of forgetting. We show that several classic sources of evidence for time-based forgetting can be re-interpreted in terms of an interference-based view, and that new experiments provide compelling evidence against decay. We conclude that progress requires moving beyond demonstrations of qualitative effects and focusing instead on testing quantitative predictions of models. [Copyright &y& Elsevier]

How do people report the contents of short-term memory when information about order must be retained but items can be retrieved in any order? We report an experiment using an unconstrained reconstruction task in which people can report list items in any order but must place them in their correct serial positions. We found (1) a tendency to report recent items first in immediate but not in delayed reconstruction, (2) a tendency to recall temporally isolated items first, (3) a preference for forward recall order, and (4) a preference for output orders that minimize the length of the path that must be traversed through memory space during retrieval. The results constrain most current models of short-term memory in which retrieval is ballistic and is assumed to run to completion autonomously once initiated. [ABSTRACT FROM AUTHOR]

Abstract: This article presents four experiments that tested predictions of SOB (Serial Order in a Box), an interference-based theory of short-term memory. Central to SOB is the concept of novelty-sensitive encoding, which holds that items are encoded to the extent that they differ from already-encoded information. On the additional assumption that distractors are encoded into memory in the same manner as list items, the theory predicts differential effects of interfering activity based on the similarity structure of distractors. Consistent with predictions, three experiments showed that overt articulation of distractors in between recalls of list items did not affect forgetting when the same distractor was repeated multiple times, whereas forgetting was observed if several different distractors were articulated within the same time span. A fourth experiment showed that the absence of forgetting with repeated articulations of the same item was not due to compensatory attentional refreshing of memory traces. The data support the notion that forgetting from short-term memory arises from interference and are difficult to reconcile with temporal decay. [Copyright &y& Elsevier]

In numerous recent studies in short-term memory, it has been established that forward serial recall is unaffected by the temporal isolation of to-be-remembered items. These findings contradict the temporal distinctiveness view of memory, which expects items that are temporally isolated from their neighbors to be more distinct and hence remembered better. To date, isolation effects have only been found with tests that do not constrain output order, such as free recall. This article reports two experiments that, for the first time, report a temporal isolation effect with forward serial recall, using a running memory task in which the end of the list is unpredictable. The results suggest that people are able to encode and use temporal information in situations in which positional information is of little value. We conclude that the overall pattern of findings concerning temporal isolation supports models of short-term memory that postulate multidimensional representations of items. [ABSTRACT FROM AUTHOR]

Three hypotheses of forgetting from immediate memory were tested: time-based decay, decreasing temporal distinctiveness, and interference. The hypotheses were represented by 3 models of serial recall: the primacy model, the SIMPLE (scale-independent memory, perception, and learning) model, and the SOB (serial order in a box) model, respectively. The models were fit to 2 experiments investigating the effect of filled delays between items at encoding or at recall. Short delays between items, filled with articulatory suppression, led to massive impairment of memory relative to a no-delay baseline. Extending the delays had little additional effect, suggesting that the passage of time alone does not cause forgetting. Adding a choice reaction task in the delay periods to block attention-based rehearsal did not change these results. The interference-based SOB fit the data best; the primacy model overpredicted the effect of lengthening delays, and SIMPLE was unable to explain the effect of delays at encoding. The authors conclude that purely temporal views of forgetting are inadequate. [ABSTRACT FROM AUTHOR]

Abstract: In short-term serial recall, similar-sounding items are remembered more poorly than items that do not sound alike. When lists mix similar and dissimilar items, performance on the dissimilar items is of considerable theoretical interest. Farrell and Lewandowsky [Farrell, S., & Lewandowsky, S. (2003). Dissimilar items benefit from phonological similarity in serial recall. Journal of Experimental Psychology: Learning, Memory, and Cognition, 29, 838–849.] recently showed that if guessing strategies are controlled, dissimilar items on mixed lists are recalled more accurately than on pure dissimilar lists, a finding that challenges several current theories of serial recall. This article presents two experiments that extend the generality of the mixed-list advantage for dissimilar items and then applies three theories of memory—the primacy model, SIMPLE, and SOB—to the data. The simulations show that the data are best explained by the SOB theory [Farrell, S. (2006). Mixed-list phonological similarity effects in delayed serial recall. Journal of Memory and Language, 55, 587-600; Farrell, S., & Lewandowsky, S. (2002). An endogenous distributed model of ordering in serial recall. Psychonomic Bulletin &Review, 9, 59–79.] which, unlike most other current theories, posits that similarity has an effect at the time of encoding. [Copyright &y& Elsevier]

Abstract: According to temporal distinctiveness models, items that are temporally isolated from their neighbors during list presentation are more distinct and thus should be recalled better. Contrary to that expectation of distinctiveness views, much recent evidence has shown that forward short-term serial recall is unaffected by temporal isolation. We report two experiments using reconstruction of order tasks that confirmed that when report order is strictly forward, temporal isolation does not benefit performance. However, both experiments also showed that when report order is unconstrained, temporal isolation does benefit performance. The differences between forward and unconstrained report were found to be independent of whether or not people can anticipate the type of test at encoding. The presence and absence of isolation effects under two different conditions, both requiring memory for order, challenges many existing theories of memory but is compatible with the idea that multiple differentially weighted types of information contribute to memory retrieval. [Copyright &y& Elsevier]

The article presents a review of the benchmark results in short-term memory (STM) research and explores the computational theories accounting for serial order memory. It also looks at the findings to determine and select the most appropriate explanatory constructs that must be instantiated in a model for a theory of STM. The authors also cite a computational model called C-SOB, and reveals that it handles existing benchmark data and diagnostic results. They conclude by identifying the novel predictions of the model.

The notion of a link between time and memory is intuitively appealing and forms the core assumption of temporal distinctiveness models. Distinctiveness models predict that items that are temporally isolated from their neighbors at presentation should be recalled better than items that are temporally crowded. By contrast, event-based theories consider time to be incidental to the processes that govern memory and such theories would not imply a temporal isolation advantage unless participants engaged in a consolidation process (e.g., rehearsal or selective encoding) that exploited the temporal structure of the list. In this report, we examine two studies that assessed the effect of temporal distinctiveness on memory, using auditory (Experiment 1) and auditory and visual (Experiment 2) presentation with unpredictably varying interitem intervals. The results show that with unpredictable intervals temporal isolation does not benefit memory, regardless of presentation modality. [ABSTRACT FROM AUTHOR]

Abstract: According to temporal distinctiveness models, items that are temporally isolated from their neighbors during list presentation are more distinct and thus should be recalled better. Event-based theories, by contrast, deny that time plays a role at encoding and predict no beneficial effect of temporal isolation, although they acknowledge that a pause after item presentation may afford extra opportunity for a consolidation process such as rehearsal or grouping. We report two experiments aimed at differentiating between the two classes of theories. The results show that neither serial recall nor probed recall benefit from temporal isolation, unless participants use pauses to group a list. Simulations of the SIMPLE model provide convergent evidence that short-term memory for serial order need not involve temporal representations. [Copyright &y& Elsevier]

According to temporal distinctiveness theories, items that are temporally isolated from their neighbours during presentation are more distinct and thus are recalled better. Event-based theories, which deny that elapsed time plays a role at encoding, explain isolation effects by assuming that temporal isolation provides extra time for rehearsal or consolidation of encoding. The two classes of theories can be differentiated by examining the symmetry of isolation effects: Event-based accounts predict that performance should be affected only by pauses following item presentation (because they allow time for rehearsal or consolidation), whereas distinctiveness predicts that items should also benefit from preceding pauses. The first experiment manipulated inter-item intervals and showed an effect of intervals following but not preceding presentation, in line with event-based accounts. The second experiment showed that the effect of following interval was abolished by articulatory suppression. The data are consistent with event-based theories but can be handled by time-based distinctiveness models if they allow for additional encoding during inter-item pauses. [ABSTRACT FROM AUTHOR]

Most models of serial recall postulate that recalled items are suppressed and thus temporarily rendered unavailable. Response suppression can explain several results, for example the small number of erroneous repetitions and people's reluctance to report repeated items. Although it is clear that response suppression is not permanent (thus permitting renewed recall of an item on the next trial), nothing is known about its time course. We report two experiments that measured the time course of response suppression with a multiple cued-retrieval response-deadline method. Emphasis was on the extent of repetition inhibition for lists that contained a repeated item. Regardless of whether presentation was rapid (Experiment 1; 150 ms/item) or slow (Experiment 2; 500 ms/item), (a) the standard pattern of repetition inhibition and erroneous repetitions occurred and (b) repetition inhibition remained constant across increasing retrieval time. This suggests that the release from response suppression is a discrete, list-wide effect rather than a continuous, gradual wearing off. The latter conclusion is consistent with the operation of the SOB model (Farrell & Lewandowsky, 2002) but not with models that postulate complete suppression with gradual wearing off. [ABSTRACT FROM AUTHOR]

Algorithms for approximate Bayesian inference, such as those based on sampling (i.e., Monte Carlo methods), provide a natural source of models of how people may deal with uncertainty with limited cognitive resources. Here, we consider the idea that individual differences in working memory capacity (WMC) may be usefully modeled in terms of the number of samples, or "particles," available to perform inference. To test this idea, we focus on two recent experiments that report positive associations between WMC and two distinct aspects of categorization performance: the ability to learn novel categories, and the ability to switch between different categorization strategies ("knowledge restructuring"). In favor of the idea of modeling WMC as a number of particles, we show that a single model can reproduce both experimental results by varying the number of particles—increasing the number of particles leads to both faster category learning and improved strategy‐switching. Furthermore, when we fit the model to individual participants, we found a positive association between WMC and best‐fit number of particles for strategy switching. However, no association between WMC and best‐fit number of particles was found for category learning. These results are discussed in the context of the general challenge of disentangling the contributions of different potential sources of behavioral variability. [ABSTRACT FROM AUTHOR]

Highlights: [•] We propose an active removal process that operates in an item-specific fashion. [•] Removal serves WM updating and general WM functioning by preventing interference. [•] WM updating involves switching between maintenance and updating modes of WM. [Copyright &y& Elsevier]

We respond to the comments of Logie and Vandierendonck to our article proposing benchmark findings for evaluating theories and models of short-term and working memory. The response focuses on the two main points of criticism: (a) Logie and Vandierendonck argue that the scope of the set of benchmarks is too narrow. We explain why findings on how working memory is used in complex cognition, findings on executive functions, and findings from neuropsychological case studies are currently not included in the benchmarks, and why findings with visual and spatial materials are less prevalent among them, (b) The critics question the usefulness of the benchmarks and their ratings for advancing theory development. We explain why selecting and rating benchmarks is important and justifiable, and acknowledge that the present selection and rating decisions are in need of continuous updating. The usefulness of the benchmarks of all ratings is also enhanced by our concomitant online posting of data for many of these benchmarks. [ABSTRACT FROM AUTHOR]

Any mature field of research in psychology--such as short-term/working memory--is characterized by a wealth of empirical findings. It is currently unrealistic to expect a theory to explain them all; theorists must satisfice with explaining a subset of findings. The aim of the present article is to make the choice of that subset less arbitrary and idiosyncratic than is current practice. We propose criteria for identifying benchmark findings that every theory in a field should be able to explain: Benchmarks should be reproducible, generalize across materials and methodological variations, and be theoretically informative. We propose a set of benchmarks for theories and computational models of short-term and working memory. The benchmarks are described in as theory-neutral a way as possible, so that they can serve as empirical common ground for competing theoretical approaches. Benchmarks are rated on three levels according to their priority for explanation. Selection and ratings of the benchmarks is based on consensus among the authors, who jointly represent a broad range of theoretical perspectives on working memory, and they are supported by a survey among other experts on working memory. The article is accompanied by a web page providing an open forum for discussion and for submitting proposals for new benchmarks; and a repository for reference data sets for each benchmark. [ABSTRACT FROM AUTHOR]

We review the evidence for the 3 principal theoretical contenders that vie to explain why and how working memory (WM) capacity is limited. We examine the possibility that capacity limitations arise from temporal decay; we examine whether they might reflect a limitation in cognitive resources; and we ask whether capacity might be limited because of mutual interference of representations in WM. We evaluate each hypothesis against a common set of findings reflecting the capacity limit: The set-size effect and its modulation by domain-specificity and heterogeneity of the memory set; the effects of unfilled retention intervals and of distractor processing in the retention interval; and the pattern of correlates of WM tests. We conclude that--at least for verbal memoranda--a decay explanation is untenable. A resource-based view remains tenable but has difficulty accommodating several findings. The interference approach has its own set of difficulties but accounts best for the set of findings, and therefore, appears to present the most promising approach for future development. [ABSTRACT FROM AUTHOR]