slides - School of Life Sciences

Document Sample
slides - School of Life Sciences Powered By Docstoc
					Memory: Working Memory
    and Forgetting
         Dr Sue Turnbull
   Clinical Neuropsychologist
 Structure and Processes
 Multi-Store Model of Memory
 Standard Modal of Short-Term Memory
 Working Memory
 Theories of Forgetting
        Structures and Processes
 Structures/Architecture
     Way in which the memory system is
     e.g. a series of stages
 Processes
     Activities occurring within the memory system
     e.g. encoding and retrieval processes
                  Multi-Store Model

Atkinson & Shiffin (1968)
               Sensory Stores
 Iconic   Store
     Transient visual store
     Information decays within 0.5 secs

 Echoic   Store
     Transient auditory store
     Information decays within 2 secs
       Short & Long-Term Stores
 William   James (1890)

     Primary memory – psychological present

     Secondary memory – psychological past
 Short-Term Store Demonstrations
 Digit   Span

 Word    List
   Short-Term Store Experiments
Extremely limited capacity
     Miller (1956) 7 ±2

 Fragile
     Glanzer & Cunitz (1966) recency effect
                     Miller (1956)
   Similar span for items containing different
    amounts of information
       Binary digits
       Decimal digits
       Letters
       Digits and letters
       Monosyllabic words from a vocabulary of 1000
   Immediate span is 7 ±2 ‘chunks’ of information
               Miller (1956)


      binary   decimal   letters   letters &   words
      digits    digits               digits

Data from Hayes (1952)
         Glanzer & Cunitz (1966)
 Fifteen   word list read out
     Immediately after learning trial write words
     10 sec delay during which subject counts out
      loud before writing words down
     30 sec delay during which subject counts out
      loud before writing words down
 Recency effect disappears with delay and
 rehearsal suppression
Glanzer & Cunitz (1966)

Proportion Correct

                     0.6                                               no delay
                                                                       10 sec delay

                                                                       30 sec delay


                           1   2 3 4   5 6   7 8 9 10 11 12 13 14 15
                                        Serial Position
                       Brown-Peterson Task
  Letter recall (%)

                            0   3   6         9         12      15   18
                                    Retention interval (secs)

Peterson & Peterson (1959)
      Evidence from Case Studies
       Dissociations Between Short- and
 Double
 Long-Term Stores
     Amnesic patients have poor long-term
      memory and intact short-term memory (e.g.
     Cases when intact long-term memory but
      impaired short-term memory
       • KF (Shallice & Warrington, 1970)
       Amnesic Patients (e.g. HM)
“Bilateralsurgical lesions in the hippocampal region…
  produce a remarkedly severe and persistent memory
  disorder. Patients seem largely incapable of adding new
  information to long-term store. This is true whether
  acquisition is measured by free recall, recognition or
  learning with savings. Nevertheless, the immediate
  registration of new input…appears to take place normally
  and material that can be encompassed by verbal
  rehearsal is held for many many minutes without further
  loss…Interruption of rehearsal… produces immediate
  forgetting of what went before… and … material that
  cannot be categorised in verbal terms decays in 30
  seconds. Material already in long-term store is
  unaffected by the lesion”
                    Milner cited by Atkinson & Shiffrin (1968)
      Amnesic Patients (e.g. HM)
“HM was able to retain the number 584 for at least 15
  minutes, by continuously working out elaborate
  mnemonic schemes. When asked how he had been able
  to remember the number for so long he replied “Its easy.
  You just remember 8. You see 5, 8 and 4 add to 17. You
  remember 8, subtract it from 17 and it leaves 9. Divide 9
  in half and you get 5 and 4, and there you are 584: Easy”
  A minute or two later, HM was unable to recall either the
  number 584 or any of the associated complex train of
  thought; in fact he did not know that he had been given a
  number to remember.”
                                          Milner (1970, p37)
 Left parieto-occipital damage
 No difficulty with long-term learning
 Digit-span impaired

                   (Shallice & Warrington, 1970)
      Support for Multi-Store Model
 Systematic  account of memory structures
 Conceptual distinctions between sensory,
  short-term and long-term stores make
 Major differences between stores
     Temporal duration
     Storage capacity
     Forgetting mechanisms differ
     Effects of brain damage
    Problems with Multi-Store Model
   Assumption that the short- and long-term stores
    are unitary
       Evidence that more than one short-term store
       Evidence that different and distinct forms of long-term
        storage systems
 Flow of information only goes one way through
  memory stores
 Role of rehearsal in transferring information to
  long-term store not applicable to everyday lives
Standard Model of Short-Term Memory

   Short-term storage = ‘activated’ information
   Relevant permanent knowledge is activated
   Short-term memory is activity trace of activated
   Activation decays over time
   Rehearsal refreshes activation and maintains
    knowledge in short-term memory
   Trade off between rehearsal and decay
  Standard Model of Short-Term Memory

                           Activation (rehearsal)

Force of Gravity = Decay

                            Lost to gravity (decayed)
      Support for Standard Model
      that are rehearsed rapidly are
 Words
 remembered better that words that are not
     Baddeley et al (1975) found a word length
      effect in that lists of words that were able to
      be spoken quickly were remembered better
      than lists of long words
     Ellis & Homelly (1980) compared digits
      presented in English (quick to pronounce) and
      digits presented in Welsh (slow to pronounce)
      and found longer digit span in English
   Problems with Standard Model
 Differences  in memory span when
 articulation rates are controlled for
     Phonologically dissimilar > Phonologically
      similar (e.g. Schweckent et al, 1990)
     Words > Nonwords (e.g. Hulme et al, 1991)
     High frequency words > low frequency words
      (Hulme et al, 1990)
     Concrete > Abstract (Walker & Hulme, 1999)
   Problems with Standard Model
 Differences  in memory span when
 articulation is suppressed
     Words were recalled better than non-words in
      a condition that prevented articulation of
      words (Caza & Belleville, 1999)
   Problems with Standard Model
 Decay   = “loss of trace information
  exclusively as a function of time” but
  forgetting is affected by factors other than
 Role of proactive interference
     Keppel & Underwood (1962)
     Tehan & Humphries (1996) (cue dependant)
     Tehan & Humphries (1996)

Block 1: jail – silk – orange – peach

Block 2: page – leap – carrot – witch

Recall the vegetable from the 2nd block
Recall the juice from the 2nd block
    Problems with Standard Model
 Forgetting rates are fixed, like gravity, but data
  shows that they are variable
 Main vehicle for short-term storage is rehearsal,
  but data shows that immediate retention can be
  independent of rehearsal
 Remembering as a direct by-product of
  activation, but data supports interpretation
  through a cue-driven retrieval process to explain
  how we remember in short-term

Nairne (2002) Annual Review of Psychology, 53, pp53-81
Baddeley’s Working Memory System

 Baddeley (2000) The episodic buffer: A new component
 of working memory? Trends in Cognitive Science, 4, 417-423
 Dual-Task Study (Robbins, 1996)
 Components of W-M in chess playing
 examined by using following tasks
     Central executive: random number generation
     Visuospatial sketchpad: pressing keys clockwise
     Phonological loop: articulatory suppression “see-saw”
     Control: repetitive tapping
 Selectingmoves involved central
 executive and the visuo-spatial sketchpad
 but not the phonological loop
             Phonological Loop
       Short-Term Verbal Memory
 Phonological store
     Passive
     Directly concerned with speech perception
 Articulatory   rehearsal system
     Articulatory process linked to speech
      production that gives access to phonological
     Support For Phonological Loop
   Experimental Studies
       Phonological similarity effect (Larsen et al, 2000)
       Word-length effect (Baddeley, 1975; Ellis & Homelly,
   Case-Studies
       Patients with damaged phonological store but intact
        articulatory control process (left tempero-parietal
       Intact phonological store + damaged articulatory
        control process
   Imaging Studies
       fMRI found left inferior parietal lobe for phonological
        store and left prefrontal for rehearsal (Henson et al,
    Challenges to Phonological Loop
   Mechanism of short-term forgetting is
       Trace Decay?
       Interference?
   How is serial order of items maintained?
       Chaining? Computational models?
   Separate phonological store or temporary
    activation of long-term memory?

Baddeley (2002) Is Working Memory Still Working?, European Psychologist, 7, pp85-97
            Visuospatial Sketchpad
Temporary storage and manipulation of spatial and
                visual information
   Visual Cache
       Stores information about visual form and colour
   Inner Scribe
       Spatial and movement information
       Rehearses information from visual cache
       Transfers information from visual cache to the central
       Involved in the planning and execution of body
                                                  Logie (1995)
Support for Visuospatial Sketchpad
   Experimental Studies
       Quinn and McConnell (1996) found tasks
        differentiating between visual and spatial aspects of
        short-term memory
   Case-Studies
       Evidence of patients with damaged visual cache and
        intact inner scribe
   Imaging Studies (e.g. Smith & Jonides, 1997)
       Visual imagery tasks – left parietal
       Spatial tasks – right hemisphere (frontal, parietal,
Challenges to Visuospatial Sketchpad

 Separating   the subcomponents of
 sketchpad has been found to be more
 difficult than for phonological loop
Central Executive (Baddeley, 1996)
 Attentionalsystem
 Located in frontal lobes
 Functions:
     Switching of retrieval plans
     Timesharing in dual-task studies
     Selective attention to certain stimuli while
      ignoring others
     Temporary activation of long-term memory
      Support for Central Executive
 Time-sharing      in dual-tasks
     Alzheimer’s patients have difficulty distributing
      attention across 2 tasks
     fMRI identified dorsolateral prefrontal cortex
      during dual-task conditions
 Selective    attention
     Stroop effect – anterior cingulate cortex
red       blue            green      tan

red       blue            green      brown

      anterior cingulate gyrus
Questions about Central Executive
       mechanism or different
 Unitary
     Individuals performance differs between tasks
      (e.g. shifting attention, updating information,
      response inhibition Miyake et al (2000)
     Brain imaging indicates disparate areas
      (prefrontal involved in all executive tasks, but
      other areas of frontal lobes plus parietal lobes
      also frequently light up)
Questions about Central Executive
         verbal and spatial working
 Separate
 memory systems?
     No correlation between ability to perform
      spatial and verbal working memory tasks
      (Shas & Miyake, 1996)
               Episodic Buffer
 Integrates information from a range of
  sources into a single complex structure or
 Acts as an intermediary between the
  phonological loop and visuo-spatial
  sketchpad combining them into a unitary
 Places high demands on attentional
  system of central executive
      Support for Episodic Buffer
 Use of verbal and visual coding in memory span
  for Arabic numerals and digit words (Chincotta et
  al, 1999)
 Increased span for words in sentences not found
  in those with impaired short-term phonological
  memory and intact long-term memory (Baddeley
  et al, 1987)
 Immediate prose recall better in amnesics with
  intact executive functioning compared to those
  with severe executive deficit (Baddeley &
  Wilson, 2002)
Baddeley’s Working Memory System

 Baddeley (2000) The episodic buffer: A new component
 of working memory? Trends in Cognitive Science, 4, 417-423
Support for Working Memory Model
 Working  memory involves active
  processing of information and transient
  storage (vs just transient storage in short-
  term store model)
 Explains partial deficits of short-term
  memory seen in brain injury (e.g. verbal,
  visuo-spatial, central executive)
 Verbal rehearsal is optional process
  within phonological loop
  Challenges for Working Memory
 Role   of central executive still unclear
     Capacity hard to measure
     Modality free?
     Composite of many components?
 Episodic   buffer
     how does it integrate information?
     Does it have a location?
 Interference   Theory

 Cue-Dependent     Forgetting

 Consolidation
             Interference Theory
 Proactive    Interference (PI) (main effect)
     Previous learning interferes with later learning

 Retroactive    Interference (RI)
     Later learning disrupts earlier learning

 Individuals   are passive to interference

 Forgettingas a direct consequence of
 acquiring and storing new memories
           Underwood (1957)
 Observed   that results from studies of
  recall over a 24 hour period varied greatly
  even if similar stimuli
 Found nearly all variability explained by
  the number of prior lists students had
 Dominance of Proactive Interference
Challenges for Interference Theory
 Can   use active processes to minimise
  interference effect
 Active inhibition can cause forgetting
 Sleep can increase retention (Jenkins &
 Dallenbach, 1924)
 Can  interference effects be found in
  implicit memory?
 Applications to real-life situations?
       Cue-Dependant Forgetting
        (1974) proposed two main ways in
 Tulving
 which we forget
     Trace-dependent – information no longer
      stored in memory
     Cue-dependent – information is still in
      memory but cannot be retrieved
       Support for Cue-Dependent
 Tulving      & Psotka (1971) no RI in cued
 Mood-state dependent memory
     Kenealy (1997)
     Memories of depressed subjects: Clark & Teasdale (1982) Journal of
      Abnormal Psychology, 91, 87-95

 Encoding specificity principle (Tulving)
 Recognition Memory > Free Recall
                                                          Kenealy (1997)

                             100                                                                       100

                             90                                                                        90

                                                                          learned items recalled (%)
learned items recalled (%)

                             80                                                                        80

                             70                                                                        70

                             60                                                                        60

                             50                                                                        50
                                   Happy                            Sad                                      Happy                           Sad
                                              mood at learning                                                         mood at learning

                                     sad at recall      happy at recall                                        sad at recall     happy at recall
      Challenges to Cue-Dependent
 The amount of information overlap does
 not always predict recall
   e.g. discriminatory ability of cue important
  (write, right, rite, write)
 Complex    recall strategies
 Intrinsic and extrinsic context have
  different effects
   Process that lasts for several hours or days and
    fixes new information in long-term memory
   Hippocampus has important role in consolidating
    newly formed memories
   Mental exertion and memory formation interfere
    with consolidation process (limited resources)
   New memories are still being consolidated and
    are vulnerable to interference and forgetting
   “New memories are clear but fragile and old
    ones are faded but robust” (Wixted, 2004)
         Support for Consolidation
 Fits   with forgetting curve
     Rate of forgetting decreases over time
 Retrograde   amnesia
 Sleep benefits learning
 Effects of alcohol on learning
     Retrograde facilitation

                                        Wixted (2004)
    Challenges to Consolidation
 Further details of the consolidation
 Other reasons that newly formed
  memories disrupted
 What about the discriminatory cues?
   Process that lasts for several hours or days and
    fixes new information in long-term memory
   Hippocampus has important role in consolidating
    newly formed memories
   Mental exertion and memory formation interfere
    with consolidation process (limited resources)
   New memories are still being consolidated and
    are vulnerable to interference and forgetting
   “New memories are clear but fragile and old
    ones are faded but robust” (Wixted, 2004)
         Support for Consolidation
 Fits   with forgetting curve
     Rate of forgetting decreases over time
 Retrograde  amnesia
 Temporal Grade of Retroactive
 Retrograde Facilitation
     Sleep benefits learning
     Effects of alcohol on learning

                                          Wixted (2004)
                  Forgetting Curve
 All forgetting functions are characterised by an
  ever decreasing perpetual rate of decay
 Fits with idea of consolidation in that as they
  consolidate they become more resistant to
  interference and therefore survive longer





          1   2   3   4   5   6   7
            Retrograde Amnesia
   Ribot’s (1881/1882) Law: after brain injury the
    most recently acquired memories are impaired
    the most
   Brain injury disrupts brain activity and therefore
    prevents consolidation from occurring
   The minutes and hours before and injury are not
    remembered as the memories have not been
   The newer the memory the more fragile it is and
    more likely to be affected
   Particularly likely if injury involves hippocampus
    (Manns et al, 2003; Squire et al, 2001)
      Temporal Grade of Retroactive
 Miller   & Pilzecher (1900)
     Interference in learning if a distracter list
      presented 17sec after, no interference if 6min
     Psychological processes associated with
      learning perseverate for a period serving to
      consolidate learning (Wixted, 2004)
          Retrograde Facilitation
 Sleep   increases recall
     Ekstrand (1972)
     Sleep decreases mental exertion and
      therefore less interference
 Improved  memory for material studied just
 prior to consumption of alcohol or
     Prevents formation of new memories that
      would otherwise cause retroactive
    Challenges to Consolidation
 Further  details of the consolidation
 Could be various reasons that newly
  formed memories disrupted
 Why do similar stimuli cause greater
 Retrograde amnesia can stretch back for
  months and years
Eysenck & Keane (2005) Cognitive Psychology: A Student’s
  Handbook, 5th Edition (ch 7)
Sternberg & Wagner (1999) Readings in Cognitive Psychology

Short-Term and Working Memory
Baddeley (2002) Is working memory still working?, European
  Psychologist, 7, pp85-97
Nairne (2002) Remembering over the short-term: The case
  against the standard model, Annual Review of Psychology, 53,

Wixted (2004) The psychology and neuroscience of forgetting,
  Annual Review of Psychology, 55, pp235-69

Shared By: