Long-term working memory in text production

In reading and other high-level cognitive tasks, Ericsson and Kintsch (1995) proposed that the limited capacity of short-term working memory (STWM) is supplemented by long- term working memory (LTWM) for individuals with a high degree of domain-specific knowledge. In Experiment 1, college students (N = 80) wrote persuasive and narrative texts concerning baseball; domain-specific knowledge about baseball and verbal ability was assessed. The results showed that verbal ability and domain-specific knowledge independently affected writing skill, supporting the view that literacy depends on both knowledge sources and refuting one argument raised in support of the LTWM hypothesis. Experiment 2 (N = 42) replicated this outcome and tested the prediction that a high degree of domain-specific knowledge would lessen interference on a secondary task. The data supported the interference prediction, offering evidence that LTWM plays a role in the production of text.     

Predicting memory benefits in the production effect: the use and misuse of self-generated distinctive cues when making judgments of learning

The production effect occurs when reading a word aloud leads to better memory for the item, relative to words that are read silently. In the present study, we assessed the degree to which judgments of learning (JOLs) are sensitive to the production effect, to determine whether people are aware of how distinctive cues can enhance memory. If the act of saying a word aloud is used as a cue for later memorability, then JOLs should be sensitive to production. Experiment 1 demonstrated that this was the case, as participants provided higher JOLs for produced items than for those read silently. This pattern of JOLs was also evident when participants silently mouthed words (Exp. 2). In Experiment 3, participants instead made a nonunique response as the production component (saying “yes” instead of the word itself). JOLs were still higher under production, although memory performance did not differ from that in a silent condition. The results suggest that the presence of both specific and nonspecific self-generated cues is used to make metacognitive judgments, likely due to the high accessibility of this information, but that participants are not precisely aware of how distinctiveness enhances encoding and retrieval. Such findings have implications for how distinctiveness is perceived by learners and for what cues would appropriately be incorporated when predicting future memory performance.     

Test-enhanced learning: taking memory tests improves long-term retention

Taking a memory test not only assesses what one knows, but also enhances later retention, a phenomenon known as the testing effect. We studied this effect with educationally relevant materials and investigated whether testing facilitates learning only because tests offer an opportunity to restudy material. In two experiments, students studied prose passages and took one or three immediate free-recall tests, without feedback, or restudied the material the same number of times as the students who received tests. Students then took a final retention test 5 min, 2 days, or 1 week later. When the final test was given after 5 min, repeated studying improved recall relative to repeated testing. However, on the delayed tests, prior testing produced substantially greater retention than studying, even though repeated studying increased students' confidence in their ability to remember the material. Testing is a powerful means of improving learning, not just assessing it.     

The drawing effect: Evidence for reliable and robust memory benefits in free recall

In 7 free-recall experiments, the benefit of creating drawings of to-be-remembered information relative to writing was examined as a mnemonic strategy. In Experiments 1 and 2, participants were presented with a list of words and were asked to either draw or write out each. Drawn words were better recalled than written. Experiments 3–5 showed that the memory boost provided by drawing could not be explained by elaborative encoding (deep level of processing, LoP), visual imagery, or picture superiority, respectively. In Experiment 6, we explored potential limitations of the drawing effect, by reducing encoding time and increasing list length. Drawing, relative to writing, still benefited memory despite these constraints. In Experiment 7, the drawing effect was significant even when encoding trial types were compared in pure lists between participants, inconsistent with a distinctiveness account. Together these experiments indicate that drawing enhances memory relative to writing, across settings, instructions, and alternate encoding strategies, both within- and between-participants, and that a deep LoP, visual imagery, or picture superiority, alone or collectively, are not sufficient to explain the observed effect. We propose that drawing improves memory by encouraging a seamless integration of semantic, visual, and motor aspects of a memory trace.     

Associative learning improves visual working memory performance

The ability to remember visual stimuli over a short delay period is limited by the small capacity of visual working memory (VWM). Here the authors investigate the role of learning in enhancing VWM. Participants saw 2 spatial arrays separated by a 1-s interval. The 2 arrays were identical except for 1 location. Participants had to detect the difference. Unknown to the participants, some spatial arrays would repeat once every dozen trials or so for up to 32 repetitions. Spatial VWM performance increased significantly when the same location changed across display repetitions, but not at all when different locations changed from one display repetition to another. The authors suggest that a major role of learning in VWM is to mediate which information gets retained, rather than to directly increase VWM capacity.     

The simultaneous acquisition effect: simultaneous task learning inhibits memory for order

Delayed recall of a list of words learned simultaneously with other lists is superior to that of a list of words learned singly. A Brown-Peterson-like task was used to investigate this simultaneous acquisition effect from the perspective of the item-order distinction. It was hypothesized that simultaneous task learning would impede the encoding of order information but promote the encoding of item-specific information relative to single-task learning. The results of the first four experiments strongly supported the hypothesis that simultaneous task learning decreases the encoding of order information but provided no evidence that it facilitated the encoding of item-specific information. Additionally, Experiment 5 showed that the simultaneous acquisition effect did not occur in a mixed-list design, demonstrating an important boundary condition of the effect.     

The effect of task relevance instructions on memory for text with seductive details

Seductive details in text are interesting but unimportant text segments. Although seductive details can make expository text more interesting, they do not necessarily promote learning of main ideas. This study investigated whether task relevance instructions that targeted main ideas would promote memory for main ideas when students read a text with seductive details. Undergraduates (n = 102) read a text with seductive detail sentences and then did a free recall task. Before reading, participants received pre‐reading questions that either targeted main ideas or seductive details, or they read for understanding. The main finding was that students in the main idea relevance instruction condition recalled main ideas better than students in the control or seductive detail relevance instruction conditions. Theoretical and practical implications of the results are discussed, and directions for future research are provided.     

The interactive effect of working memory and text difficulty on metacomprehension accuracy

We investigated how working memory capacity (WMC) and text difficulty affect metacomprehension accuracy. Participants completed the operation-span test to measure WMC and ability to read expository texts. Under the easy-text condition, participants read 4 texts with increasing local cohesion, whereas under the difficult-text condition, participants read 4 original texts. Participants assigned a comprehension rating to each text and then completed a comprehension test. The results revealed a significant interactive effect of WMC and text difficulty on metacomprehension accuracy in Experiments 1 and 2. Under the easy-text condition, higher-WMC readers monitored their comprehension less accurately than did lower-WMC readers. In contrast, higher-WMC readers monitored their comprehension more accurately than did lower-WMC readers under the difficult-text condition. These results suggest that text difficulty may affect allocation of attentional resources.     

The costs and benefits of testing text materials

Tests have been shown to improve the later recall of tested information, a result known as the testing effect. Tests, however, can also impair the later recall of related information, an effect known as retrieval-induced forgetting. Although retrieval-induced forgetting has been demonstrated using a wide variety of materials, recent work suggests that learning information in the context of a coherent text passage may afford protection from retrieval-induced forgetting. In four experiments we explored the conditions under which retrieval-induced forgetting does and does not occur with such materials. We found that two factors—the coherence of the to-be-learned material and the competitiveness of retrieval practice—are important in determining whether retrieval-induced forgetting does or does not occur. Furthermore, even if retrieval-induced forgetting does occur, having the opportunity to restudy the forgotten information can prevent that forgetting from persisting. Taken together, these findings provide greater understanding of the costs and benefits of testing text materials, with possible implications for the optimisation of testing as a tool for learning in educational contexts.     

The costs and benefits of testing text materials

Tests have been shown to improve the later recall of tested information, a result known as the testing effect. Tests, however, can also impair the later recall of related information, an effect known as retrieval-induced forgetting. Although retrieval-induced forgetting has been demonstrated using a wide variety of materials, recent work suggests that learning information in the context of a coherent text passage may afford protection from retrieval-induced forgetting. In four experiments we explored the conditions under which retrieval-induced forgetting does and does not occur with such materials. We found that two factors-the coherence of the to-be-learned material and the competitiveness of retrieval practice-are important in determining whether retrieval-induced forgetting does or does not occur. Furthermore, even if retrieval-induced forgetting does occur, having the opportunity to restudy the forgotten information can prevent that forgetting from persisting. Taken together, these findings provide greater understanding of the costs and benefits of testing text materials, with possible implications for the optimisation of testing as a tool for learning in educational contexts.     

The production effect benefits performance in between-subject designs: A meta-analysis

Producing (e.g., saying, mouthing) some items and silently reading others has been shown to result in a reliable advantage favoring retention of the produced compared to non-produced items at test. However, evidence has been mixed as to whether the benefits of production are limited to within- as opposed to between-subject designs. It has even been suggested that the within-subjects nature of the production effect may be one of its defining characteristics. Meta-analytic techniques were applied to evaluate this claim. Findings indicated a moderate effect of production on recognition memory when varied between-subjects (g = 0.37). This outcome suggests that the production effect is not defined as an exclusively within-subject occurrence.     

The production effect in paired-associate learning: Benefits for item and associative information

In five experiments, we extended the production effect—better memory for items said aloud than for items read silently—to paired-associate learning, the goal being to explore whether production enhances associative information in addition to enhancing item information. In Experiments 1 and 2, we used a semantic-relatedness task in addition to the production manipulation and found no evidence of a production effect, whether the measure was cued recall or item recognition. Experiment 3 showed that the semantic-relatedness task had overshadowed the production effect; the effect was present when the semantic-relatedness task was removed, again whether cued recall or item recognition was the measure. Experiments 4 and 5 provided further evidence that production can enhance recall for word pairs and, using an associate recognition test with intact versus rearranged pairs, indicated that production may also enhance associative information. That production boosts memory for both types of information is considered in terms of distinctive encoding.     

Trial-and-error learning improves source memory among young and older adults

Trial-and-error learning, relative to errorless learning, has been shown to impair memory among older adults, despite evidence from young adults that errors may afford memorial benefits through richer encoding. However, previous studies on the effects of errorless versus trial-and-error learning in older adults has required production of errors based on perceptual cues. We hypothesized that producing errors conceptually associated with targets would boost memory for the encoding context in which information was studied, especially for older adults who do not spontaneously elaborate on targets at encoding. We report two studies examining the impact of generating errors during learning on source memory among young and older adults, with a process dissociation procedure employed in Study 1, and source memory assessed directly in Study 2. In both studies, participants were shown semantic category cues and generated an exemplar either with or without errors. In Study 1, for both age groups trial-and-error learning was associated with lower familiarity-based memory and higher recollection-based memory relative to errorless learning, and the latter effect was more marked for older than younger adults. Similarly, in Study 2, trial-and-error learning was associated with better source memory relative to errorless learning, particularly for the older adults. We argue that trial-and-error learning can enhance source memory and confer memorial benefits when making such errors facilitates semantic elaboration, especially for older adults who do not spontaneously engage in strategic encoding.     

Sleep directly following learning benefits consolidation of spatial associative memory

The last decade has brought forth convincing evidence for a role of sleep in non-declarative memory. A similar function of sleep in episodic memory is supported by various correlational studies, but direct evidence is limited. Here we show that cued recall of face-location associations is significantly higher following a 12-h retention interval containing sleep than following an equally long period of waking. Furthermore, retention is significantly higher over a 24-h sleep-wake interval than over an equally long wake-sleep interval. This difference occurs because retention during sleep was significantly better when sleep followed learning directly, rather than after a day of waking. These data demonstrate a beneficial effect of sleep on memory that cannot be explained solely as a consequence of reduced interference. Rather, our findings suggest a competitive consolidation process, in which the fate of a memory depends, at least in part, on its relative stability at sleep onset: Strong memories tend to be preserved, while weaker memories erode still further. An important aspect of memory consolidation may thus result from the removal of irrelevant memory "debris."     

The effect of layout and pacing on learning from diagrams with unnecessary text

Although the presentation of extraneous (i.e., irrelevant or unnecessary) information hinders learning, it is unclear whether and how layout and pacing influence this effect. In two experiments, participants learned how the heart functions using four different layouts: a diagram presented without unnecessary text (diagram only), with unnecessary text separated from the diagram (separated) or integrated into the diagram (integrated), or with separated unnecessary text and the instruction to integrate (integration instruction). In Experiment 1, study time was self‐paced for half of the participants and system paced for the other half. There were no effects of layout and of pacing on learning, although system pacing was more effortful than self‐pacing. In Experiment 2, which was system paced and employed eye tracking, the integrated condition showed worse learning outcomes than the separated condition. Moreover, in the integrated condition, participants made more integration attempts between the unnecessary text and the diagram than in the separated condition.     

Working memory span: the effect of prior learning

Recent work suggests that working memory span (WMS) tasks are not simple measures of the capacity to simultaneously store and process new information. Instead, these measures may be influenced by numerous factors, including proactive interference (PI). The current study examined whether WMS, like other memory tasks, is also influenced by PI from prior memory experiments. Experimentally experienced and naive participants completed a speaking span task. Span scores were lower for experienced than for naive participants, but other cognitive scores were not. In combination with other work, these results suggest that WMS estimates are not pure measures of capacity and may be partially determined by PI.     

The costs and benefits of memory conformity

We examined the influence of external recommendations on memory attributions. In two experiments, participants were led to believe that they were viewing the responses of two prior students to the same memoranda they were currently judging. However, they were not informed of the reliability of these fictive sources of cues or provided with performance feedback as testing proceeded. Experiment 1 demonstrated improvement in the presence of reliable source cues (75% valid), as compared to uncued recognition, whereas performance was unaltered in the presence of random cues provided by an unreliable source (50% valid). Critically, participants did not ignore the unreliable source, but instead appeared to restrict cue use from both sources to low-confidence trials on which internal evidence was highly unreliable. Experiment 2 demonstrated that participants continued to treat an unreliable source as potentially informative even when it was predominantly incorrect (25% valid), highlighting severe limitations in the ability to adequately discount unreliable or deceptive sources of memory cues. Thus, under anonymous source conditions, observers appear to use a low-confidence outsourcing strategy, wherein they restrict reliance on external cues to situations of low confidence.     

Distributed rereading can hurt the spacing effect in text memory

The spacing effect refers to the commonly observed phenomenon that memory for spaced repetitions is better than for massed repetitions. In the present study, we examined this effect in students' memory for a lengthy expository text. Participants read the text twice, either in immediate succession (massed repetition), with a 4‐day interstudy interval (spaced short), or with a 3.5‐weeks interstudy interval (spaced long). Two days after the second study trial, all participants were tested. The results demonstrated that students in the spaced‐short condition remembered more of the content than those in the massed condition. By contrast, students in the spaced‐long condition remembered as much as students in the massed condition. These results were interpreted in terms of a theoretical framework, which combines mechanisms of encoding variability and study‐phase retrieval to account for the spacing effect. Copyright © 2007 John Wiley & Sons, Ltd.     

Age differences in learning from text: Evidence for functionally distinct text processing systems

We investigated the influence of sentence elaboration on self‐regulated learning in order to examine age differences in resource allocation to the construction of textbase‐ and discourse‐level representations. Older and younger adults learned about a topic by reading a series of sentences varying in elaboration (from simple factoids to highly elaborated text) and manner of presentation (progressive change in elaboration vs. random change in elaboration). Younger readers were more likely to recall information from factoids; older adults, from highly elaborated text. Relative to young, older readers showed an advantage in the progressive presentation condition, which minimised frequent changes between textbase‐ and discourse‐level processing. Older adults showed poorer memory monitoring for factoids and less elaborated discourse relative to young, but when passages were highly elaborated or presented progressively, age differences were eliminated. Results support the idea that textbase‐ and discourse‐level encoding arise from functionally distinct systems whose regulation depends on text properties and reader age. Copyright © 2008 John Wiley & Sons, Ltd.