Prospective memory: The relative effects of encoding,retrieval, and the match between encoding and retrieval

Previous prospective memory studies have revealed some important features of encoding, retrieval, and the match between the encoding and the retrieval that contribute to prospective memory performance. However, these studies have not provided evidence concerning the relative importance of these three factors because no study has investigated all three in a single design. We developed a laboratory-based paradigm that allowed us to manipulate different characteristics of encoding, retrieval, and the match between encoding and retrieval simultaneously in a single experiment. The results of eight experiments showed that all three factors have an influence on prospective memory performance, but that the match between encoding and retrieval has a significantly larger influence than either encoding or retrieval factors.     

Prospective memory: the relative effects of encoding, retrieval, and the match between encoding and retrieval

Previous prospective memory studies have revealed some important features of encoding, retrieval, and the match between the encoding and the retrieval that contribute to prospective memory performance. However, these studies have not provided evidence concerning the relative importance of these three factors because no study has investigated all three in a single design. We developed a laboratory-based paradigm that allowed us to manipulate different characteristics of encoding, retrieval, and the match between encoding and retrieval simultaneously in a single experiment. The results of eight experiments showed that all three factors have an influence on prospective memory performance, but that the match between encoding and retrieval has a significantly larger influence than either encoding or retrieval factors.     

Concurrent task effects on memory encoding and retrieval: Further support for an asymmetry

Several studies have demonstrated that divided attention at encoding significantly reduces memory performance, whereas divided attention at retrieval affects memory performance only minimally. However, the possibility exists that retrieval processes have shown such resilience because the concurrent tasks used have often not been very demanding. To assess this possibility, we used independent manipulations of the concurrent task during either encoding or retrieval that included stimulus-response compatibility and participant- versus experimenter-controlled pace. In addition, we manipulated the distribution of practice that the participants received with the primary and the concurrent tasks. The results replicated and extended those recently reported by Rohrer and Pashler (2003), indicating that although memory performance is negatively affected by divided attention at retrieval, especially with noncompatible stimulus-response mapping in the concurrent task, this effect was much smaller than that at encoding, in line with the asymmetry notion. Furthermore, experimenter versus participant control of the concurrent task had no effect on memory retrieval. Finally, under conditions of equal practice with both the memory and the concurrent tasks, memory retrieval was affected only to a small degree. In contrast to encoding processes, the processes involved in retrieval accuracy appear, in many cases, to be less interrupted by divided attention, although this protection requires substantial resources.     

Levels of processing, encoding strategies, and memory development.

In three experiments, memory for intentionally encoded words was compared with memory for encodings induced by asking semantic, phonemic, or surface questions. Subjects were second-, fourth-, and sixth-grade, and junior and senior high school students. Semantic encodings were more often recalled and recognized than were phonemic and surface (which did not differ). Intentional encodings were as likely to be recalled as semantic but were no more likely than phonemic and surface to be recognized, and this was true whether intentional learning was in anticipation of a recall test or a recognition test. Age trends occurred for recall, for intentional learning, and for induced processing which required subjects to generate word attributes. Age trends were attenuated for recognition and for induced processing which required subjects to verify whether a give attribute applied to the target word.     

Effects of correspondence between encoding and retrieval organization in social memory

Memory research has shown impaired recall performance when a subset of the studied stimuli is presented at recall (the part-list cueing effect, Slamecka, 1968) or when the recall is collaborative (collaborative-inhibition effect, Weldon & Bellinger, 1997). In two experiments we explore these effects in an impression-formation context and compare two prominent accounts (retrieval blocking versus strategy disruption) for them. We varied the correspondence between item organization at encoding and retrieval, either by manipulating the organization of part-list cues (Experiment 1) or the organization of the stimulus list that was later recalled collaboratively (Experiment 2). Results showed that when encoding and recall organizations did not correspond recall was impaired, replicating part-list cueing and collaborative-inhibition effects. However, when encoding and retrieval organization corresponded, these effects were greatly reduced. Such results support the recall strategy disruption hypothesis and challenge the retrieval blocking account. Implications for understanding memory in a social context are discussed.     

An activation-based model of sentence processing as skilled memory retrieval

We present a detailed process theory of the moment-by-moment working-memory retrievals and associated control structure that subserve sentence comprehension. The theory is derived from the application of independently motivated principles of memory and cognitive skill to the specialized task of sentence parsing. The resulting theory construes sentence processing as a series of skilled associative memory retrievals modulated by similarity-based interference and fluctuating activation. The cognitive principles are formalized in computational form in the Adaptive Control of Thought–Rational (ACT–R) architecture, and our process model is realized in ACT–R. We present the results of 6 sets of simulations: 5 simulation sets provide quantitative accounts of the effects of length and structural interference on both unambiguous and garden-path structures. A final simulation set provides a graded taxonomy of double center embeddings ranging from relatively easy to extremely difficult. The explanation of center-embedding difficulty is a novel one that derives from the model' complete reliance on discriminating retrieval cues in the absence of an explicit representation of serial order information. All fits were obtained with only 1 free scaling parameter fixed across the simulations; all other parameters were ACT–R defaults. The modeling results support the hypothesis that fluctuating activation and similarity-based interference are the key factors shaping working memory in sentence processing. We contrast the theory and empirical predictions with several related accounts of sentence-processing complexity.     

Neurophysiology of successful encoding and retrieval of source memory

Event-related potentials were recorded during encoding, to identify whether brain activity predicts subsequent retrieval of spatial source information, and during retrieval, to investigate the neural correlates of successful and unsuccessful spatial context recollection. The amplitude registered during encoding for study items that were later recognized and assigned a correct source judgment was more positive than the amplitude for recognized items given incorrect source judgments; this difference started 480 msec poststimulus, predominantly at central and anterior sites. During retrieval, the waveform was more positive from 250 to 1,600 msec poststimulus when the brain had retrieved episodic information successfully than when it had failed. These findings indicate that brain electrical activity recorded during the first presentation of an item within its context predicts the subsequent retrieval of the specific spatial context. During retrieval, brain activity differed quantitatively at anterior sites and qualitatively at posterior sites according to the accuracy of source memory.     

记忆编码中目标与背景的情绪一致性对记忆提取的影响

摘 要 本研究将经典情绪启动范式和学习-再认范式相结合,探讨学习时启动与目标的情绪一致性对目标再认的影响。在记忆编码阶段,以情绪面孔图片为启动刺激、一般情绪图片为目标刺激,要求被试对目标图片的效价进行判断;在提取阶段向被试呈现一般情绪图片,要求被试判断是否在前一阶段见过。结果表明,编码阶段的加工背景对记忆提取产生了显著影响。一致启动促进了认知加工与记忆编码,提高了再认正确率。ERP结果也表明,在目标刺激呈现后300-500ms,编码时的情绪一致性对再认阶段消极目标诱发的ERP产生了显著影响。     

Distinguishing the neural correlates of episodic memory encoding and semantic memory retrieval

Episodic memory and semantic memory interact very closely. In particular, episodic memory encoding (EE) tends to elicit semantic memory retrieval (SR), and vice versa. Thus, similar activations for EE and SR in functional neuroimaging studies may reflect shared memory processes, or they may reflect the fact that EE and SR are usually confounded. To address this issue, we used a factorial functional magnetic resonance imaging approach to disentangle the neural correlates of EE and SR. Within the left temporal lobe, the hippocampus was associated with successful EE, whereas a posterior lateral region was associated with successful SR. Within the left inferior prefrontal cortex, a posterior region was involved in SR, a mid region was involved in both SR and EE, and an anterior region was involved in EE, but only when SR was also high. Thus, the neural correlates of EE and SR are dissociable but interact in specific brain regions.     

Generation,intentionality of processing at encoding and retrieval,and age-related associative deficits

This research tested age-related differences in the retrieval of self-generated new associations under conditions that required intentional or incidental processing. Under intentional or incidental encoding conditions, young and older adults generated new associations by producing a response to a two-letter stem paired with a cue/prime word (e.g., throne–mo_____). Memory for these new associations was tested under intentional or incidental retrieval conditions by pairing the word stem with the previous cue/prime word, its homophone partner, or a prime/cue not previously presented. Results indicated equivalent priming and cueing effects for both age groups in all conditions. These results suggest that generation of new associations can eliminate age-related associative deficits, even under intentional encoding and retrieval conditions that typically disadvantage older adults.     

Associative asymmetry,availability, and retrieval

Associative frequency, the ease with which a word comes to mind in free association, is taken as a measure of general response availability. As expected from this view, in both controlled experiments and in reanalyses of previously published correlational data, high associative frequency words were judged to be more familiar and were easier to recall but harder to recognize than low associative frequency words, even with meaningfulness, imagery, length in letters, and frequency excluded as factors. When used as foils in a recognition experiment, high associative frequency words attracted more responses than low associative frequency words. In addition, associative frequency and meaningfulness correlated only moderately and had different patterns of correlations with other variables, suggesting that the number of associations leading to and from a word differ.     

Neural processing associated with true and false memory retrieval

We investigated the neural bases for false memory with fMRI by examining neural activity during retrieval processes that yielded true or false memories. We used a reality monitoring paradigm in which participants saw or imagined pictures of concrete objects. (A subsequent misinformation task was also used to increase false memory rates.) At test, fMRI data were collected as the participants determined whether they had seen or had only imagined the object at study. True memories were of seen pictures accurately endorsed as seen, and for false memories were of imagined pictures falsely endorsed as seen. Three distinct patterns of activity were observed: Left frontal and parietal activity was not different for true and for false memories, whereas activity was greater for true than for false memories in occipital visual regions and posterior portions of the parahippocampal gyrus, and activity was greater for false than for true memories in right anterior cingulate gyrus. Possible interpretations are discussed.     

Prefrontal and hippocampal contributions to encoding and retrieval of spatial memory

The prefrontal cortex is thought to be critical for goal-directed action and the hippocampus is known to be importantly involved in spatial memory. Several studies have been suggestive of a role for the orbitofrontal cortex (OFC) in spatial navigation. However, the medial prefrontal cortex (mPFC) receives projections directly from the intermediate CA1 (iCA1) region of hippocampus and this link may be critical for spatial navigation. The purpose of the present investigation was to test the performance of rats receiving bilateral or disconnection infusions of lidocaine into OFC, mPFC, or iCA1 to determine the contribution of these structures to encoding and retrieval of spatial memory using the Hebb–Williams maze. A total of 92 male Long-Evans rats received chronic bilateral, contralateral, or ipsilateral implantation of cannulas into OFC, mPFC, or iCA1. Prior to testing on day 1 or day 2, subjects received central infusions of saline or lidocaine. The number of errors committed on the first five trials compared to the second five trials of day 1 was used to determine encoding, whereas retrieval was determined by comparing the second five trials of day 1 with the first five trials of day 2. The present findings suggest that mPFC and iCA1 are necessary and interact during encoding and retrieval; however, the OFC does not appear to be essential for either process. While the nature of the interaction between mPFC and iCA1 during encoding and retrieval is unclear, it may be supported by the integration of goals and spatial cues or strategy switching.     

Interactive effects in transfer-appropriate processing for event-based prospective memory: The roles of effort, ongoing task, and PM cue properties

Past studies (e.g., Marsh, Hicks, & Cook Journal of Experimental Psychology: Learning, Memory, and Cognition 31:68–75, 2005; Meiser & Schult European Journal of Cognitive Psychology 20:290–311, 2008) have shown that transfer-appropriate processing (TAP) effects in event-based prospective memory (PM) depend on the effort directed toward the ongoing task. In the present study, we addressed mixed findings from these studies and examined monitoring in TAP and transfer-inappropriate processing (TIP) conditions. In two experiments, a semantic or orthographic ongoing task was paired with a PM cue that either was matched in processing (TAP) or did not match in processing (TIP). Within each condition, effort was varied across trials. The results indicated that PM accuracy was higher in TAP than in TIP conditions, regardless of effort condition, supporting the findings reported by Meiser and Schult. Ex-Gaussian functions were fit to the mean reaction times (cf. Brewer Journal of Psychology 219:117–124, 2011) in order to examine monitoring across conditions. The analysis of distributional skew (τ parameter) showed sensitivity to ongoing task instructions and properties of the PM cues. These results support Meiser and Schult’s suggestion that TIP conditions require more attentional processing, and they also afford novel discussion on the interactive effects of ongoing task condition, PM cue properties, and manipulations of effort.     

Relations between emotion,memory encoding,and time perception

This study examined duration judgments for taboo and neutral words in prospective and retrospective timing tasks. In the prospective task, participants attended to time from the beginning and generated shorter duration estimates for taboo than neutral words and for words that they subsequently recalled in a surprise free recall task. These findings suggested that memory encoding took priority over estimating durations, directing attention away from time and causing better recall but shorter perceived durations for taboo than neutral words. However, in the retrospective task, participants only judged durations in a surprise test at the end, and their duration estimates were longer for taboo than neutral words. Present findings therefore suggest that the same emotion-linked memory encoding processes can cause underestimation of durations in prospective tasks but overestimation in retrospective tasks, as if emotion enhances recall of ongoing events but causes overestimation of the durations of those events in retrospect.     

Context reinstatement and memory for intrinsic versus extrinsic context: The role of item generation at encoding or retrieval

According to many theoretical accounts, reinstating study context at the time of test creates optimal circumstances for item retrieval. The role of context reinstatement was tested in reference to context memory in several experiments. On the encoding phase, participants were presented with words printed in two different font colors (intrinsic context) or two different sides of the computer screen (extrinsic context). At test, the context was reinstated or changed and participants were asked to recognize words and recollect their study context. Moreover, a read‐generate manipulation was introduced at encoding and retrieval, which was intended to influence the relative salience of item and context information. The results showed that context reinstatement had no effect on memory for extrinsic context but affected memory for intrinsic context when the item was generated at encoding and read at test. These results supported the hypothesis that context information is reconstructed at retrieval only when context was poorly encoded at study.