Serial Position Effect

10 Key excerpts on "Serial Position Effect"

• 

  eBook - ePub

  Comparative Perspectives on the Development of Memory

  • R. V. Kail, Jr., N. E. Spear(Authors)
  • 2013(Publication Date)
  • Psychology Press

    (Publisher)

  In summary, the spacing effect is a facilitation of relatively long-term retention due to a relatively long interval between learning trials. The effect characterizes performance of rats, monkeys, apes, human infants, and adults on several memory tasks. There is some question whether birds such as pigeons can be included in this list. The spacing effect has been demonstrated in paradigms involving classical and operant conditioning, habituation of orienting responses and reflexes, and verbal learning.

  Serial Position EffectS

  Discussions of Serial Position Effects inevitably address the context in which learning and retention occur. This is because Serial Position Effects happen when an item (word, picture, or event) to be remembered appears in a sequence of other items (a list). Additionally, the list itself is in a context of extralist events, and the proximity of an item to this extralist context may be important. As noted by Ebbinghaus (1885/1917), acquisition and forgetting of an item are related to its serial position in the list—if the item occurred in the middle of the sequence, it is less likely to be remembered than if it occurred at the beginning or end. The benefit of position at the beginning of a list is termed “the primacy effect” and the benefit of position at the end of a list is termed “the recency effect.” Like spacing effects, Serial Position Effects are abundant in experimental and naturalistic studies of verbal mnemonics (see review by Crowder, 1976, chapter 12 ). The advantages for recall of the extreme positions in a list are so common as to be indicated in the time necessary to answer the following question: What day of the week is today? The response on Wednesday takes longest, and midweek names are also more prone to confusion than Sunday or Friday (Koriat & Fischoff, 1974). Those of us who live for the weekends may not be surprised at this result, even though recall of other semantic knowledge is not found to be attenuated when assessed at midweek (Koriat & Fischoff, 1974).

  Serial Position Effects can be systematically revealed in a probe recognition memory test. The general features of this procedure are illustrated in a study of habituation of an orienting response in rabbits (Wagner & Pfautz, 1978). Five distinctive stimuli were ordered to make different lists for different subjects. The stimuli were short (1 second) sensory events: a tone, a click, a flashing light, a vibrating massage, or a faint electric current. The interstimulus interval was 2 seconds and, because 8 or 16 exposures of the list were necessary to produce habituation, the interlist intervals were 150 seconds. The orienting response was vasoconstriction, an autonomic reaction to reduce blood flow to the extremities prior to fright or fight. To the extent the animal is familiar with a harmless event, there is habituation of such vasomotor activity. Retention of habituation was probed after separate repetitions of an invariant sequence of sensory events. In this case, the probe followed the last list presentation by 15 minutes, a retention interval void of stimulus manipulations. The probe consisted of 12 presentations of one of the stimuli (the 1 second vibration) that had consistently occupied either the first, middle, or last position in the list. The probe indicated the amplitude of vasoconstrictive orienting was high if the vibration had occupied the middle position in the list, but appreciably lower if the vibration had occurred at the beginning or conclusion of the list. The analysis indicated retention of habituation was a nonmonotonic function of serial position; primacy and recency effects were apparent at the extremes of a bowed plot of peak response (see Fig. 12.4

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Human Memory

  Paradigms and Paradoxes

  • Robert L. Greene(Author)
  • 2014(Publication Date)
  • Psychology Press

    (Publisher)

  recency effect refers to the advantage at the end. Although this would depend in part on how the experiment was done, the recency effect in immediate free recall is usually somewhat larger than the primacy effect.

  The Primacy Effect

  The topic of this chapter is not primacy effects. Rather, this chapter will focus on recency effects. This concentration is consistent with the interests of researchers, who in the last 30 years at least have paid much more attention to the recency effect than to the primacy effect. This relative lack of interest reflects the feeling that psychologists have achieved a satisfactory (and unfortunately somewhat uninteresting) account of primacy effects. This account assumes that early items are remembered better than middle or late items simply because they receive more rehearsal or attention.

  FIG. 3.1 . An example of the recency effect in immediate free recall.

  Imagine that you are a conscientious subject receiving a list of items one at a time that you are trying to memorize. When the first item occurs, it has your undivided attention. You rehearse it in some way until the next item appears. When the second item appears, you try to rehearse it, but you also try to keep the first item accessible by rehearsing it as well. In other words, you would begin sharing rehearsals between the current item and the first item. As more items are presented, this process of sharing rehearsal becomes more difficult. Since there are more and more previous items to try to maintain, you give less rehearsal to the current item as you get further into the list. In other words, since you are always dividing your cognitive effort between the current item and earlier ones, the first few items would end up with a greater cumulative amount of rehearsal or processing than later items. The primacy effect could then reflect the fact that items are easier to remember if they received a lot of rehearsal or processing.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Principles of Learning and Memory

  Classic Edition

  • Robert G. Crowder(Author)
  • 2014(Publication Date)
  • Psychology Press

    (Publisher)

  The two-process model for position effects in single-trial situations are just about as unlikely in multitrial serial learning. At least, attribution of recency to an echoic store or to a fragile primary memory buffer is almost unthinkable in the context of a subject clamped into a rigid, paced anticipation schedule where each item must emerge in its fixed 2-sec slot. There is nothing wrong, however, with application of the primacy part of the two-process model to serial learning; that is, there could be active, selective processing working to the advantage of the early items provided the pacing of the rate of presentation were not too frantic.

  Therefore, it should be noted, two hypotheses emerge relatively unscathed from this effort to apply theories of serial-position effects in multitrial situations to single-trial situations and vice versa: The idea that serial-position effects come from the greater distinctiveness, somehow, of the end items and the idea that primacy comes from an active processing schedule that favors early items. This point is important because we shall return to conclude that primacy probably comes from the differential processing at the beginning of the list and recency from the differential distinctiveness at the end of the list. However, if we take seriously our responsibility to formulate an account of serial-position effects that generalizes among all or most of the situations that yield a serial-position effect, there are several more groups of data to consider. These data further limit the hypotheses that we can apply consistently.

  The conceptual position effect . The ordinary assumption among students of verbal learning brought up in the tradition of Ebbinghaus has been that, at least in the laboratory, verbal items may be considered relatively independent of one another; furthermore, when this is not the case, workers in this tradition have probably wished it were. Since the earliest stirrings of the field of psycholinguistics, there has been another research tradition called verbal behavior

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Essential Cognitive Psychology

  • Alan J. Parkin(Author)
  • 2014(Publication Date)
  • Psychology Press

    (Publisher)

  In the early 1960s the serial position curve was reinterpreted in terms of James’ distinction between primary and secondary memory. The high recall of most-recently presented items was termed the recency effect and thought to reflect the effortless output of primary memory. Recall from the rest of the list was attributed to recall from secondary memory, with enhanced recall of the first few items being, somewhat confusingly, known as the primacy effect, and the middle named the asymptote (Fig. 4.2). The serial position curve was not considered sufficient proof of the distinction between primary and secondary memory. What was required was the demonstration of a functional double dissociation. The logic of this is as follows: If different parts of the curve reflect the operation of different forms of memory, and those different forms of memory presumably have different properties, then it should be possible to find factors that affect one section(s) of the curve but not the other. Specifically, different factors should influence recall from the recency items compared with those in the primacy and asymptote positions. Early experiments had already shown that certain variables raised or lowered the amount recalled from primacy and asymptote positions but had little effect on recency. Lepley (1934), for example, found that practising memory for nonsense syllables greatly increased memory at early list positions but had no appreciable effect on recency. Subsequent experiments identified a number of other experimental factors that had exactly the same kind of influence and these are summarised in Fig. 4.3a–d. FIG. 4.3. (a)–(d) Examples of experimental manipulations that enhance recall from early list positions but do not affect recency

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Learning and Memory

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

    (Publisher)

  In experiments on free recall, we’ve seen that participants are given lists of words to read and then allowed to recall them in any order they want. You might think that participants would be equally good at remembering words from all parts of the list, but this turns out not to be the case. If the probability of recalling a word is plotted as a function of the word’s position in the list – this is known as a serial position curve – the result usually resembles the curve shown in Figure 8.9 . As shown in the figure, the probability of recall is greater for words from the beginning of the list than for words in the middle, and greater still for words from the end. The heightened recall of words from the beginning of the list is known as the primacy effect, and the improved recall of words from the end is called the recency effect. The obvious question is why recall should be greater for words from the beginning and end of the list. Let’s consider first how the model accounts for the recency effect. If participants are asked to recall a list immediately after they finish reading it, the last words in the list will still be in STS. All participants have to do is to recall these words first, while they are still relatively fresh in STS, and they will obtain high scores for the words at the end of the list. Moreover, if they begin by recalling the last 312 Memory Recall primacy recency 16 1 Serial position in list Figure 8.9 A typical serial position curve, showing the likelihood of recalling a word from a list as a function of its position in the list. word and work backward, then earlier words will continue to decay as they say the later ones, thus accounting for the fact that the last word is recalled best, then the next-to-last word, and so on. As predicted by this analysis, participants do in fact begin by recalling the words from the end of the list first (Welch and Burnett, 1924 ). The model explains the primacy effect largely by rehearsal.

  Sign up to read

  Learn more about book
• 

  eBook - PDF

  Fundamentals of Learning and Memory

  • John P. Houston(Author)
  • 2014(Publication Date)
  • Academic Press

    (Publisher)

  This fact is called the p r i m a c y effect. Items in the middle of the list are the most difficult to recall. This pattern is quite stable and has been obtained by many investigators (e.g., M u r d o c k , 1962; Tulving & Arbuc-kle, 1963). The Primacy Effect Let us first dispense with the primacy effect, w h i c h is of m i n o r interest to us in attempting to distinguish short-term f r o m long-term memory. The general assumption c o n c e r n i n g the primacy effect is that it re-flects a greater a m o u n t of rehearsal given to early items in the list. Early items are recalled fairly well because they are rehearsed more than later items (see Rundus, 1971). Put yourself in the position of rehearsing a serial list. One item is exposed at a time. W h e n the first item appears, you rehearse it over and over again as many times as you can before the second item appears. W h e n the second item does appear, you probably rehearse from the b e g i n n i n g of the list, repeating items one and t w o over and over again. W h e n item three appears, you Figure 12.4 Idealized serial position curve for free recall. 1.0 h .20 I I I Serial position 370 ISSUES IN MEMORY still attempt to rehearse f r o m the b e g i n n i n g of the list, and so o n . If this rehearsal strategy is f o l l o w e d , it is clear that early items will be rehearsed m o r e than later items. Hence, they are learned and recalled better than later items (see also Reynolds & H o u s t o n , 1964). The Recency Effect Why are the last items recalled best of all? The rehearsal strategy just outlined cannot a c c o u n t for the recency effect. The most c o m m o n interpretation of the recency effect (see Glanzer, 1972; Kintsch & Pol-son, 1979; M u r d o c k , 1974) involves the distinction between short-term and long-term m e m o r y . During presentation, all items enter short-term m e m o r y . As additional items are presented, s o m e of the earlier items are lost.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Essential Cognitive Psychology (Classic Edition)

  • Alan J. Parkin(Author)
  • 2013(Publication Date)
  • Psychology Press

    (Publisher)

  negative recency effect: words that were remembered well initially, i.e. those comprising the recency effect, were subsequently remembered very poorly relative to items presented earlier (Craik, 1970).

  FIGURE 4.3

  (a)–(d) Examples of experimental manipulations that enhance recall from early list positions but do not affect recency. (e)The effect of distraction prior to recall affects recency, but not earlier list positions (adapted from Crowder, 1976)

  Short-term and long-term store

  During the 1960s the information processing concept began to dominate experimental psychology. The model that had the most obvious effect on memory research was Atkinson and Shiffrin’s multistore model (Atkinson and Shiffrin, 1971). This model is shown in Figure 4.4 and characterises memory as a series of stores between which various forms of information flow. External information first enters sensory registers which are able to store it briefly before it then enters short-term store (STS). This is equivalent to James’ primary memory in that STS supports conscious mental activity. STS was able to apply various control processes to the information it contained. These processes included: rehearsal—essentially a recycling mechanism functionally equivalent to the idea of repeating things to yourself; coding—the extraction of certain features from new information to form the basis of a new memory; retrieval strategies—mechanisms for accessing information stored in long-term store (LTS; secondary memory); and decision processes

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  The Foundations of Remembering

  Essays in Honor of Henry L. Roediger, III

  • James S. Nairne(Author)
  • 2011(Publication Date)
  • Taylor & Francis

    (Publisher)

  Figure 12.5 , contrary to that account, there was no effect of condition and, in fact, no effect of serial position. The three conditions show very different functions, but even the interaction of condition and serial position was not significant, presumably because of inconsistencies within as well as across conditions.

  Summary

  In sum, the reconstruction of order data are inconsistent with the general predictions based on the distinctiveness account as well as the specific predictions based on Murdock’s (1960) formulation. The primacy effects predicted on the basis of Murdock’s formulation are not at all evident. There were, however, recency effects in two of the conditions, but the recency effect was strong only in the 25–42 condition, where the recent items were the highly familiar living presidents. Further, in the analysis of the overlapping positions, there was no advantage for being in the first or last parts of the list. In contrast, the data are generally consistent with the previous account in terms of familiarity (Healy et al., 2000). The strongest support for that account is from the two significant correlations we found between the functions for the reconstruction of order task and the familiarity ratings.

  EXPERIMENT 4

  Experiment 3 involved long-term semantic memory because the ordering of US presidents should reside in the subjects’ mental encyclopedia and because the temporal context of learning the presidents’ positions is not relevant. We wondered whether the distinctiveness of the relative or absolute serial positions might play a larger role in immediate episodic memory, where bow-shaped serial position functions were originally demonstrated by Nipher (1878) and others. To examine this issue, in Experiment 4 we turned to an immediate episodic memory task for actors (see Healy et al., 2000, for a similar manipulation). This experiment was designed to be as closely analogous as possible to Experiment 3 except that instead of reconstructing the order of US presidents from semantic memory, subjects reconstructed the order of famous actors from immediate episodic memory. Episodic memory was involved in this case because the ordering of the actors is based on when the subjects encountered each actor’s name during the experiment.

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Attention and Performance VI

  Proceedings of the Sixth International Symposium on Attention and Performance, Stockholm, Sweden, July 28–August 1, 1975

  • Stanislav Dornic̆(Author)
  • 2022(Publication Date)
  • Routledge

    (Publisher)

  Note that the recency effect is equally marked on immediate test in all three list lengths, but that after only 15 seconds of interpolated activity it has virtually disappeared, while performance on earlier items is relatively unimpaired. This result has been interpreted in a range of different ways, but by far the most common and influential has been the suggestion that the recency effect represents the contents of a labile short-term store, while performance on earlier items is based on a more durable long-term storage system (Atkinson & Shiffrin, 1968 ; Baddeley, 1968 a; Glanzer & Cunitz, 1966). Other interpretations include the suggestion that the forgetting of the most recent items occurs because of proactive interference (PI) from earlier items (Postman & Phillips, 1965), that the recency effect represents the output of an articulatory-rehearsal buffer (Atkinson & Shiffrin, 1968), that the recency effect represents a change in the subject’s strategy (Craik & Lockhart, 1972), and that it represents the application of a specific retrieval strategy (Tulving, 1968). Fig. 1 Serial position curve in recall as a function of length of list and retention interval. (After Postman & Phillips, 1965.) One way of tackling this issue might be to spell out all the various theories that have been suggested, together with the vast amount of evidence that has been produced over the last decade, and attempt to synthesize this into an adequate overall theory of recency. However, a good deal of the material has already been summarized by Shallice (1975), Watkins (1974), and Craik and Jacoby (1975). A further review would be both redundant and distinctly indigestible, and would not, alas, allow one to come up with a final convincing solution. What we propose to do instead is to present three groups of experiments that have substantially modified our own view of recency

  Sign up to read

  Learn more about book
• 

  eBook - ePub

  Interactions Between Short-Term and Long-Term Memory in the Verbal Domain

  • Annabel Thorn, Mike Page(Authors)
  • 2008(Publication Date)
  • Psychology Press

    (Publisher)

  3 The roles of short-term and long-term verbal memory in free and serial recall: Towards a recency-based perspective

  Geoff Ward, Lydia Tan and Parveen Bhatarah

  Overview

  In this chapter, we reinterpret classic empirical evidence that has been argued previously to support a theoretical distinction between short-term and long-term memory stores (STS and LTS, respectively) within a recency-based framework of episodic memory. Our discussion centres on the free recall and immediate serial recall tasks, from which, historically, measures of the recency effect and memory span respectively have provided some of the most important lines of empirical evidence in establishing the dichotomy.

  Free recall

  In the free recall task, participants are presented with a sequence of unrelated items for study, one at a time, and immediately after the presentation of the last item, they must try to remember as many of the list items as possible, freely recalling the items in any order that they wish. Typically, free recall gives rise to U-shaped or J-shaped serial position curves, which show that the early items and the later items in the list tend to be recalled more often than the middle list items, and these advantages are known as the primacy effect and the recency effect, respectively (e.g., Murdock, 1962).

  The classic interpretation of the serial position curve is that performance on the early and middle list items is due to retrieval from LTS, whereas performance on the later list items is due to retrieval from STS and LTS. Within this account, the primacy effect is assumed to reflect the increased associative strength of the early items in LTS that results from additional rehearsals afforded to these list items (Atkinson & Shiffrin, 1968, 1971; Rundus, 1971), whereas the recency effect is assumed to be the result of the direct output of items residing in STS at the time of test (e.g., Atkinson & Shiffrin, 1971; Glanzer, 1972).

  Sign up to read

  Learn more about book