Cue-Dependent Forgetting

12 Key excerpts on "Cue-Dependent Forgetting"

• 

  eBook - PDF

  Psychology

  Modules for Active Learning

  • Dennis Coon, John Mitterer, Tanya Martini, , Dennis Coon, John Mitterer, Tanya Martini, (Authors)
  • 2021(Publication Date)
  • Cengage Learning EMEA

    (Publisher)

  “Why, I remember it as clearly as if it were yesterday,” he will say, forgetting that the story he is about to tell is one he told earlier the same day (twice). In short, disuse offers no more than a partial explanation of long-term forgetting. Retrieval Failure If encoding failure and storage failure don’t fully explain for-getting from long-term memory, what does? If you have en-coded and stored information, that leaves retrieval failure as a likely cause of forgetting (Della Sala, 2010; Guerin et al., 2012). Even if memories are available (stored in your memory), you still have to be able to access them (locate or retrieve them) in order to remember. For example, as we mentioned earlier, you might have had the experience of knowing that you know the answer to an exam question (you knew it was available) but being unable to retrieve it during the exam (it was inaccessible). Cue-Dependent Forgetting One reason why retrieval may fail is that retrieval cues are missing when the time comes to access the information. For instance, if you were asked, “What were you doing on Monday afternoon of the third week in May, two years ago?” your reply might be, “Come on, how should I know?” However, if you were re-minded, “That was the day the courthouse burned” or “That was the day Stacy had her automobile accident,” you might remember immediately. The presence of appropriate cues almost always enhances memory retrieval. As we saw previously, more elaborately encoded memories are more likely to be remem-bered because more retrieval cues are associated with any particular piece of information. Memory will even tend to be better if you study in the same room where you will be tested. Because this is often impossible, when you study, try to visualize the room where you will be tested. Doing so can enhance memory later (Jerabek & Standing, 1992).

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Theoretical Aspects of Memory

  Volume 2

  • Michael Gruneberg, Peter E Morris, Michael Gruneberg, Peter E Morris(Authors)
  • 2006(Publication Date)
  • Routledge

    (Publisher)

  Mood-dependent memory must be distinguished from mood- congruent memory (e.g. Blaney 1987). Mood congruence is the finding that emotionally laden stimuli are learned and remembered best when their affective valence matches the subject’s mood. For example, in a depressed mood, a subject might learn and remember negatively valenced words, such as ‘funeral’ or ‘sorrow’, better than positively valenced words, such as ‘funny’ or ‘carnival’. Although mood congruent effects resemble mooddependent memory effects, they are not the same. Mooddependence is a principle that relates study and test moods to each other, whereas mood congruence describes a relation between subjective moods and target stimuli.

  THEORETICAL PRINCIPLES

  A number of basic principles can be induced from the empirical studies reviewed above. Although these principles are intended to explain contextual dependence, they do not necessarily imply specific theoretical mechanisms. Principles such as cue overload, outshining or decontextualization can be implemented with different theoretical mechanisms in different theories. Any memory theory, however, that involves context-dependent memory should incorporate these basic principles.

  The principles to be considered here will include cuedependence, overshadowing, contextual fluctuation, cue overload, memory probe, outshining and decontextualization.

  Cue-dependent memory

  The principle of cue-dependent memory is simply that performance on memory tasks is influenced by associated memory cues. If contextual information is associated with target material, then contextual cues should stimulate memory for associated material.

  There are many theoretical mechanisms that can explain context-dependent memory, and a few will be described here. These include activation of a set of information in memory, direct context-to-item associations, mediation by internal states, and activation of cognitive operations.

  Activation of a search set

  Shiffrin (1970) described retrieval as a probabilistic iterative process involving sampling-with-replacement from a delimited set of information in memory. The delimited set was referred to as a ‘search set’, or the set of information in memory that was to be deliberately searched. By keeping memory searches within this set, one’s retrieval efforts could be more efficient than if all of memory were searched.

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Learning and Memory

  • David A. Lieberman(Author)
  • 2011(Publication Date)
  • Cambridge University Press

    (Publisher)

  3 Retrieval depends on whether the context during retrieval again generates this particular meaning or code. In our example, if jam is encoded as a preserve, then retrieval cues will help us to recall it only if they also lead us to think of preserves. More broadly, the principle of encoding specificity emphasizes yet again the importance of contextual cues in memory. To understand retrieval, we need to take into account the context during both coding and retrieval: The greater the similarity of these contexts, the greater the chances of success. 2 2 Other theorists have also emphasized the importance of the relationship between the conditions during coding and retrieval. One version is called transfer-appropriate processing ; it focuses more on the cognitive processes carried out during coding and retrieval rather than on the external 418 Memory Inhibition We have examined how retrieval cues help us to locate material lodged in the vast storehouse that is our memory, but so far we have ignored a rather large puzzle implicit in our account, arising from the fact that retrieval cues are often associated with a very large number of targets. If you read the word cat , it is probably associated with a very, very large number of memories – facts you learned about cats (their love of fish, enjoyment of stroking, hunting of birds), interactions with individual cats you have known, and so on. When you read the word cat , how is it that you are not overwhelmed by a flood of such memories? One possible answer, developed most vigorously by Michael Anderson, is inhi-bition . His idea, with roots back to Pavlov and beyond, is that our brains rely on inhibitory as well as excitatory processes in order to function effectively. If a retrieval cue elicits multiple memories, there will be a competition between them to determine which will rise up into consciousness, a cognitive version of survival of the fittest.

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Memory

  • Elizabeth Ligon Bjork, Robert A. Bjork(Authors)
  • 1996(Publication Date)
  • Academic Press

    (Publisher)

  One could find occasional men- tion of its importance by experimental psychologists (K6hler, 1947; Melton, Memory Copyright 9 1996 by Academic Press, Inc. All rights of reproduction in any form reserved. 197 198 Henry L. Roediger, III and Melissa J. Guynn 1963), but investigators rarely created conditions to study retrieval direct- ly. An unspoken assumption guiding much research was that responses produced on a memory test indicated the contents of storage. Failures to produce information were ascribed to problems in encoding or storage. For example, forgotten information might not have been transferred from a short-term store to a longrterm store (Waugh & Norman, 1965), or it may not have been processed to a deep, semantic level (Craik & Lock- hart, 1972), or over time memory traces may have decayed or been over- written. Tulving (1974) referred to theories, such as the original levels of process- ing ideas, as trace-dependent theories of memory: the critical determinant of performance on memory tests was thought to be the status of the memory trace at the time of testing. Little attention was paid to the idea that encod- ing and storage could have succeeded~the memory trace of the experience was robust--and that forgetting could still occur due to retrieval failure. The alternative to trace-dependent forgetting theories are those embodying the assumption of Cue-Dependent Forgetting. Cue-Dependent Forgetting the- orists maintain that memory for an event is always a product of informa- tion from two sources, the memory trace and the retrieval cue, the latter being the information present in the individual's cognitive environment when retrieval occurs (Tulving, 1974, p. 74). The key idea is that both the traces of past experience and the information or cues in the cognitive system during the test are critical determinants of remembering.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Tutorial Essays in Psychology

  Volume 2

  • N. S. Sutherland(Author)
  • 2014(Publication Date)
  • Psychology Press

    (Publisher)

  4 Analyzing Memory by Cuing: Intrinsic and Extrinsic Knowledge

  Gregory V. Jones University of Oxford

  This chapter is concerned with the processes which underlie the everyday use of human memory. One of the most striking aspects of memory is the richness of the interconnections between different items of stored information. Until quite recently, however, most experimental studies of human memory have investigated the recall of isolated pieces of information such as sets of unrelated words. It can be argued (Neisser, 1976) that one of the most pressing needs within experimental psychology as a whole is to investigate processes of a degree of complexity comparable to those that occur in everyday life, and several such studies have recently been reported (e.g., Linton, 1975; Robinson, 1976).

  It must be recognized, however, that in general the price of investigating more complex examples of recall is an increase in the difficulty of maintaining experimental control. For example, in an interesting study, Robinson (1976) presented subjects with cues such as "window," "run," or "happy," and asked each person to think of a specific incident of which the word reminded him or her. One of the findings was that affect terms such as "happy" induced recall of more recent events than did object or activity terms such as "window" or "run." One interpretation might be that memories of such incidents decay particularly rapidly, and thus only comparatively recent emotional experiences are available for recall. It could be, however, that people experience incidents relating to "happy" more frequently than those relating to "window" or "run," and hence come across such incidents earlier when scanning backwards in time. To decide between these two interpretations it would be necessary to undertake an experiment in which the frequency of occurrence of such incidents was controlled. The general problem with the investigation of memory for everyday events is that there is a large range of admissible theoretical interpretations, because we do not have a sufficiently detailed knowledge of the relevant conditions.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Principles of Memory

  • Aimée M. Surprenant, Ian Neath(Authors)
  • 2013(Publication Date)
  • Psychology Press

    (Publisher)

  Models of immediate and working memory also use cues and are also cue driven. For example, in the Burgess and Hitch (1999) model, an instantiation of the phonological loop component of working memory (Baddeley, 1986), a slowly varying context-time signal is used to initiate the retrieval process. This type of cue is also used by OSCAR (Brown, Preece, & Hulme, 2000), a very different model that addresses many aspects of immediate serial recall. Yet another account of similar data is offered by the feature model (Nairne, 1990), which also requires cues. Connectionist models of memory are also cue driven (cf. Ratcliff, 1992). Humphreys, Bain, and Pike (1989) present a model that specifies the cues used not only in episodic and semantic tasks, but also those that tap procedural memory. In fact, it is hard to conceive of a computer model of memory that is not cue dependent, simply because the computer requires something to start the retrieval process.

  3.2 Memory Without Cues

  There is one area of memory research in which the opposite assumption is made, that is, memory does not require cues. For example, Kintsch, Healy, Hegarty, Pennington, and Salthouse (1999) summarized 10 different theoretical accounts of working memory, highlighting similarities and differences among the various conceptions. They explicitly noted that because of the common assumption that “information ‘in’ working memory is directly and effortlessly retrievable, … retrieval from ST-WM is not cue dependent” (pp. 413–414).2 Similarly, Wickens, Moody, and Dow (1981, p. 17) state that “the retrieval act is not required in the PM [primary memory] situation.” More recently, Jonides et al. (2008, p. 202) note that “the central point of agreement” among current models of short-term memory is that “representations are directly accessible and available for cognitive action.”

  One particular version of working memory that posits cueless retrieval is called the embedded processes model (Cowan, 1995, Cowan, 1999). Essentially, memory is divided into three parts (see Figure 3.1 ): Long-term store, activated memory, and the focus of attention. Together, the activated part of long-term memory as well as the contents in the focus of attention comprise working memory.

  According to this view, an item becomes activated when it receives attention, and activated items can enter the focus of attention when they are the object of current processing. Thus, in terms of a typical memory experiment, retrieval means “entering the correct items into the focus of attention” (Cowan, 1999, p. 75). If an item is in the focus of attention, it does not have to be retrieved. The focus of attention is limited to approximately four unrelated items (Cowan, 2000); in addition to this item-based limit,

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Cognition

  The Thinking Animal

  • Daniel T. Willingham, Cedar Riener(Authors)
  • 2019(Publication Date)
  • Cambridge University Press

    (Publisher)

  A crucial factor in sensitivity is the match between encoding and retrieval, but prior knowledge also has an effect, just as it does at encoding. In fact, prior knowledge can influence retrieval to such an extent that a completely false memory can be created. It might seem that after information is stored into long-term memory, you should be able to retrieve it whenever you want. As you know, memory doesn’ t work this way. Sometimes you try to remember who played Dumbledore in the Harry Potter movies and it pops right out of the storehouse, so to speak. Other times, you can’ t quite get it, but when someone provides the name you immediately recognize it as correct (and you confidently reject Ian McKellen and Ed Harris as incorrect) (see Figure 8.1). There are different ways to retrieve memories – or to measure whether a person remembers something – and the way memory is measured has a big impact on whether a piece of information appears to be in long-term memory. Measures of Memory Before we can talk about the details of retrieval, we need to be more precise about the different ways to measure memory. First, we need to define a cue, which is information in the environment that is used as a starting point for retrieval. If we simply say to you, “Remember, ” the command makes no sense. Are you supposed to be remembering something about pickles, your second-grade teacher, or the structure of barium? A cue for what is to be retrieved from long-term memory might be provided by the experimenter (“Try to remember what I told you an hour ago”) or the environment (an advertisement for motor oil reminds you to get your oil changed), or you might provide it yourself (you mentally retrace your steps in an effort to remember where you might have left something). Memory tests differ in the cues that the experimenter provides. We will discuss three types here: free recall, cued recall, and recognition tests.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Current Issues in Cognitive Processes

  The Tulane Flowerree Symposia on Cognition

  • Chizuko Izawa(Author)
  • 2014(Publication Date)
  • Psychology Press

    (Publisher)

  As we noted at the outset, context-dependent forgetting is a primitive assumption in many current memory models. This assumption is not often submitted to critical analysis, even though the value of such memory models as explanations of remembering and forgetting is largely predicated on its truth. The literature reviewed here suggests that our reliance on context change as the major factor explaining forgetting needs to be seriously questioned. We have argued that the range of situations under which changing incidental EC will negatively affect performance (i.e., produce forgetting) is limited, owing to the roles of outshining and mental reinstatement. Tying general explanations of forgetting to changes in incidental EC would therefore not appear to be advisable, even if EC cues are encoded and later influence retrieval under some conditions. It might be argued that changes in integral and influential context are primarily responsible for forgetting; however, this argument lacks plausibility. It is clear that large amounts of forgetting occur in situations in which no obvious changes in integral or influential context have occurred. For example, an item on a recognition test, unless highly ambiguous in meaning, is likely to be interpreted semantically in the same way as it was interpreted at study. Yet a subject might still fail to recognize that item. Change in physiological context is another often-invoked candidate in explanations of forgetting. Again, the idea that this type of context is subject to changes within an experimental session that are extreme enough to produce forgetting is simply implausible.

  Based on the evidence reviewed above concerning the role of contextual variation in reducing interference, it seems that theoretical explanations of forgetting need to focus not on contextual change, but on the interference that occurs between items that are encoded in similar contexts. In other words, we need to develop theories that deal in a systematic manner with the problems of interference that preoccupied researchers in an earlier era. As proposed by those earlier researchers (e.g., Bilodeau & Schlosberg, 1951; McGovern, 1964), contextual factors probably play an important role in modulating interference processes. However, as noted by Bjork (1989), in the brain-metaphor-influenced theoretical environment of the late 1980s, it is perhaps finally appropriate to reintroduce the additional notion of active retrieval inhibition into our explanations of forgetting and remembering.

  INCIDENTAL EC CUES IN PERSPECTIVE: TOWARD A TAXONOMY

  We have restricted our focus here to contexts of a particular type (incidental environmental context) and to effects on memory measured in a particular way (recall). In terms of the possible types of contextual features that might be reinstated, however, and in terms of the possible measures of memory that might show an influence of such reinstatement, that restriction is severe. In this final section, in an attempt to lend some organization to the entire space defined by the various types of contexts and types of measures, we propose a three-dimensional taxonomy.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Memory

  Phenomena, Experiment and Theory

  • Alan Parkin(Author)
  • 2016(Publication Date)
  • Taylor & Francis

    (Publisher)

  Generation-recognition theory proposes that retrieval involves two stages. Recall also involves these two stages, but recognition proceeds without generation. Experiments on encoding specificity suggest, by contrast, that recall and recognition reflect a common retrieval process. The debate as to which of these theories is correct continues. Experiments on recognition memory indicate that two independent processes can contribute to recognition: context retrieval and familiarity. Context retrieval is dependent on explicit recollection, whereas the processes underlying familiarity-based responding bear a considerable similarity to those responsible for implicit memory phenomena. Context-retrieval and familiarity-based recognition are associated with different experiential awareness.

  Context Effects

  Regardless of which particular model of retrieval eventually proves correct, it is clear that context plays an important role in our everyday memories. I remember watching a travel programme on TV in which the director focused on what he thought was a typical inhabitant of the Loire Valley, who was, in fact, one of my departmental colleagues. Given this unlikely context, it took me about 30 seconds to carry out a recognition response that would normally occur almost instantly.

  There are many formal demonstrations that context exerts powerful effects on our ability to identify things. In a classic experiment, Light and Carter-Sobell (1970) demonstrated how changes in the intrinsic context associated with a target could reduce recognition memory. Subjects studied simple sentences which biased subjects towards encoding one particular meaning of an ambiguous word (e.g. They were stuck in a traffic jam’). Retention testing involved subjects identifying the target words, which were again embedded in biasing sentences but, in half these, the sentence was biased toward a different meaning from that used at encoding (e.g. ‘They enjoyed eating the jam’). It was found that recognition of the targets was significantly reduced when the biasing context at test was different from that at learning.

  Disruptive effects of changing intrinsic context between learning and test are perhaps not surprising, because, in effect, what is being shown is that subjects are worse at remembering a different stimulus from the one they studied. More interesting, and potentially more relevant, is the possibility that extrinsic context could exert effects on remembering.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Current Issues in Applied Memory Research

  • Graham M. Davies, Daniel B. Wright(Authors)
  • 2009(Publication Date)
  • Psychology Press

    (Publisher)

  Thus, the notion of destructive updating where memory is kept up to date by overwriting older or out-of-date information with new (cf. Loftus, Miller, & Burns, 1978) seems a potentially costly enterprise – not only in terms of the social gaffes we are likely to make (e.g., phoning up our old flames by mistake) but also in terms of the cognitive effort and time that would be needed to relearn material (e.g., the telephone number of one’s current partner). Rather, it would seem that there is a need for memory updating to be achieved in a much more subtle and flexible manner – one in which memories that are deemed irrelevant for a current task can be excised from conscious inspection but not irrevocably lost (E. L. Bjork, Bjork, & MacLeod, 2006; M. D. MacLeod & Macrae, 2001). Thus, related but unwanted memories can be temporarily forgotten for as long as it takes to complete the current task but subsequently become available for some future task should they be required. We believe that the mechanism underlying retrieval-induced forgetting (i.e., inhibition) may offer just this kind of flexibility in memory updating.

  Retrieval practice and memory performance

  Retrieval-induced forgetting refers to a particular pattern of forgetting that occurs as a result of the selective retrieval practice of other related material. Unlike directed forgetting (e.g., Johnson, 1994; C. MacLeod, 1989), this specific form of forgetting occurs implicitly as a result of the retrieval process itself (i.e., there are no explicit instructions to forget material). Rather, forgetting occurs as a function of the retrieval competition emanating from related memories. In order to deal with unwanted competition at retrieval, related memories are actively inhibited or suppressed, thereby promoting the retrieval of the material we wish to remember.

  The standard procedure used to explore retrieval-induced forgetting typically comprises a four-phase retrieval practice paradigm (M. C. Anderson et al., 1994). Although there are a number of variations, the basic paradigm generally involves the presentation of a series of category–exemplar pairs (e.g., sport–football, sport–tennis, sport–rugby, sport–cricket . . . tree– sycamore, tree–oak, tree–spruce, tree–poplar . . . bird–sparrow, bird–eagle, bird–robin, bird–mallard . . . vegetable–carrot, vegetable–onion, vegetable– celery, vegetable–broccoli . . .). On completion, participants are cued to retrieve half of the exemplars from half of the categories. For example, participants might be required to complete the following cued stem tests (e.g., sport–fo____, sport–te____; bird–sp____, bird–ea____) but are not prompted to retrieve exemplars from the remaining categories (i.e., exemplars from tree or vegetable categories). Following a distractor task (between 5 and 20 minutes), participants are asked to recall all the exemplars that had originally been presented.

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Thinking About Human Memory

  • Michael S. Humphreys, Kerry A. Chalmers(Authors)
  • 2016(Publication Date)
  • Cambridge University Press

    (Publisher)

  In this model, if a pair of words AB is studied in a context X, a memory trace is stored that contains features from the context and the two words. At retrieval, if A is provided as a cue, features from A and the context X are simultaneously matched against all of the traces in memory. The trace that produces the largest match is then examined. The first stage of the process is the same whether A is being used to recall B or is being recognized. As acknowledged by Tulving (1983) there is a further need for processes that select B for recall and make a decision for recognition. Although nobody has proposed how these response selection processes would operate in this model it seems unlikely that they would introduce a greater level of sensitivity to context for recall than for recognition. The REM model (Shiffrin & Steyvers, 1997) also uses cues in similar ways for recogni- tion and recall, as does the Matrix model (Humphreys, Bain, & Pike, 1989). Finally, the Reder et al. (2000) model does assume that recognition will be less sensitive to context than recall. This occurs because an acontextual form of familiarity (Jacoby & Dallas, 1981; Mandler, 1980; Yonelinas, 2002) is used as well as contextual information. However, in this model, list discrimination 26 The importance of thinking about cues and targets would be as sensitive as recall, so it is not the specificity of the cue that is responsible for the reduced sensitivity of recognition to context. Thus cue specificity, as in the difference between recognition and cued recall, is not a theoretically neutral concept that can readily be mapped onto the sensitivity of context. In fact it seems unlikely that any model that truly tried to address the complexities of the episodic–semantic distinction could avoid assuming that something other than cue specificity (e.g., the information reinstated by reminding the participants of the study episode) would be required.

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Concise Learning and Memory

  The Editor's Selection

  • (Author)
  • 2010(Publication Date)
  • Academic Press

    (Publisher)

  That is, reduced spontaneous recovery has been observed when a retrieval cue for the mem-ory of extinction treatment is presented before testing following a long retention interval (Brooks and Bouton, 1993). As we previously discussed, counterconditioning is another example of the context dependency of memory. If counterconditioning training is con-ducted with an appetitive reinforcer given in one context and an aversive reinforcer given in a different context, conditioned responding to the CS is guided by the context in which subjects are tested (Peck and Bouton, 1990). The retrieval model outlined earlier simply states that whichever memory representation is facilitated by contextual cues will be more prone to guide behavior. In summary, we have reviewed general models of memory that emphasize retrieval mechanisms as crit-ical for behavioral control. In general, Spear’s (1978) model emphasizes the similarity between the total information presented during training and that pre-sented at test and by this simple principle explains memory phenomena such as state-dependent learn-ing and the Kamin effect. The comparator hypothesis (Miller and Matzel, 1988) is an associative model that emphasizes competition between representations and explains cue-competition phenomena and recovery from cue competition that does not involve further training with the target cue. Bouton’s model (1993, 1997) emphasizes the role of retrieval cues in situa-tions in which one cue has more than one meaning, as in the cases of extinction, latent inhibition, and counterconditioning. One obvious conclusion is that each model has been designed to account for a family of phenomena at the expense of explaining other phenomena. For example, the comparator hypothesis (Miller and Matzel, 1988) accounts for cue competi-tion phenomena and several other effects in classical conditioning, but it does not explain the recovery from extinction or counterconditioning effects that are consistently observed.

  Sign up to read

  Learn more about book