Memory

Question 1

Measuring memory (4 types of task)

Answer 1

• Conscious behavioural methods:
  
  • objective (accuracy/RT)
  • subjective (confidence, ratings)
• Unconscious behavioural methods
  
  • priming
  • conditioning
  • habits (past affects future)
• Physiological methods (SCR, GSR, HR)
• Electrophysiological methods (EEG, MEG)
• Haemodynamic methods (PET, fMRI)

Question 2

Multi-store model (picture) - outline parts

Answer 2

Atkinson & Shiffrin (1968)

Question 3

Sensory stores (multi-store model):

• Iconic store
• Auditory

Answer 3

Sensory stores - different store for different modality (modality specific)

• typically unaware of it unless we attend to it (limited capacity)

Iconic store (George Sperling, 1960): visual

• visual array (e.g. 4x4 digits) presented for 50ms –> could report 4 or 5
• partial report condition (repeat specific row)
  
  • could report most letters on row
• rapid timeframe –> array and prompt have to be <1s apart or there is decay

Auditory:

• Triesman (1964) –> dichotic listening
  
  • if second message starts <2s after first - recognise that they’re the same
  • unattended information stays in echoic store for ~2s

Question 4

Short-term memory (multi-store model)

• what is it?
• capacity
• rehearsal
• forgetting

Answer 4

Limited capacity (~7 ) - often unreliable, currently held in mind

Capacity:

• George Miller (1956): investigated span of STM - digit string recall
  
  • ~7+- 2 digits OR 7 chunks
  • Simon (1974) - digit span decreases as chunk size increases

Rehearsal:

• information can be retained in STM by rehearsing it
• longer it is held in ST store, greater chance of LT storage (A+S, 1968)
• Rundus (1971) - rehearse list of 20 words –> more rehearsed = more likely to be recalled
  
  • recency effect - last word in list always had high likelihood of recall

STM forgetting

• Peterson & Peterson (1959) - forgetting curve - recall decreased as time interval increased (with no rehearsal)
• Waugh & Norman (1965) - STM forgetting due to decay or interference?
  
  • if decay - fast presentation would increase recall (less chance to decay)
  • if interference - fast = slow presentation recall
  • found fast = slow presentation recall –> so interference due to new info not passage of time

Question 5

Short term vs long term stores

• difference?
• recency + primacy effect definition
• Neuropsychological studies (proving separate)

Answer 5

William James (1890): primary memory = psychological present (in consciousness); secondary memory = psychological past (has left consciousness)

Recency effect = remember last few items better (STM)

Primacy effect = more resources available for initial items (LTM)

Neuropsychological studies –> STM + LTM distinct:

• Shallice 1988 –> double dissociation = patient 1 has A not B; patient 2 has B not A
• Scoville & Milner (1957) - HM, medial temporal damage, LTM impaired, intact digit span
  
  • preserved recency effect (STM fine)
  • no primacy effect (LTM impaired)
• Shallice & Warrington (1970) - KF, parieto-occipital damage, intact LTM, poor digit span
  
  • preserved primacy effect (LTM fine)
  • no recency effect (STM impaired)

Question 6

Criticisms of the multi-store model:

• 3 assumptions and their weaknesses

Answer 6

1. Strictly serial progression

• STM required for LTM encoding
• BUT: not the case in patient KF (Shallice & Warrington, 1970)

2. Longer item is held in STM storage, greater likelihood of entering LTM storage

• rehersing items in ST store does correlate with LT retention (Rundus, 1971)
• BUT: other factors more important - such as depth processing (Craik & Tulving, 1975)

3. ST + LT stores are unitary and operate in a single uniform way

• there is a single ST store and a single LT store
• BUT: Warrington & Shallice (1972) - KF had worse STM for auditory stimuli than visual
• BUT: Baddeley & Hitch (1974) - dual task methodology
  
  • if concurrent performance decreases assuracy or increases RT - there is interference so they must tap into same cognitive systems
  • auditory reheardal of digits does not increase errors made in concurrent grammatical reasoning task

Question 7

Addition to multi-store model: working memory

Answer 7

Baddeley & Hitch (1974):

• Phonological loop - storage system for auditory/visual information
• Visuo-spatial sketchpad - analogous system for non-auditory/visual information
• Central executive - modality-free processing system - coordinates operations of others

Baddeley (2000):

• Added episodic buffer –> limited capacity system that holds + integrates diverse information

Question 8

Evidence for Baddeley & Hitch’s working memory model - dual task methodology

Answer 8

Robbins et al., 1996:

• play chess while doing three tasks that tap into different aspects of WM:
  
  • rapid word repetition (phonological loop)
  • sequential key pressing (visuospatial sketchpad)
  • random number generation (need executive processing)
• Chess performance worse for 2+3 (not 1) –> visuospatial + central executive involved in chess

Question 9

Phonological loop (WM):

• phonological similarity effect
• word length effect
• phonological store vs articulatory control process

Answer 9

Phonological similarity effect (Baddeley, 1966)

• serial recall of list of phonologically similar words worse than phonologically dissimilar words
• visual + semantic similarity has little effect on recall
• so we must use speech-based representations when storing words:
  
  • phonological memory traces

Word length effect (Baddeley et al., 1975)

• serial recall significantly worse for longer words
• the lower the digit speech duration (e.g. in Chinese vs English), the higher mean digit span (number of digits remembered)
• Articulatory suppression during presentation and recall eliminates word length effect
  
  • so capacity of phonological loop is determined by how much you can rehearse it

Phonological store vs articulatory processes (Baddeley, 1990)

• phonological store = speech perception; articulatory processes = speech production
• if presented with word auditorily - goes straight into phonological store (rapid)
• if presented with word visually - goes into loop via articulatory control process (not automatic)

Question 10

Visuospatial sketchpad (WM):

• Behavioural evidence
  
  • different processes involved
  • different components of it
• Neuropsychological data

Answer 10

VS = temporary storage and manipulation of visual and spatial information

Behavioural data:

• Baddeley et al., 1975: encode info via rote learning or imagery-based strategy
  
  • task combined with pursuit rote tracking (a visuospatial task)
  • imagery-based strategy was disrupted but not rote learning
• Baddeley & Lieberman (1980) - rote vs imagery task again
  
  • combined with brightness judgements (visual) or judging position of pendulum auditorily (spatial)
  • imagery-based task was disrupted during spatial task
• Logie (1995) - split visuospatial sketchpad into: visual cache and inner scribe
  
  • visual cache = passive store of visual info. - subject to decay and interference
  • inner scribe = processes spatial info. + allows for active rehearsal of info in visual cache

Neuropsychological data:

• Beschin et al., (1997) - NL - preserved perceptual skills but couldn’t describe scene details from memory
• Farah et al., (1988) - LH –> better at spatial processing tasksthan visual imagery tasks

Question 11

Episodic buffer (WM)

• more than just modality-specificity of VS + PL
• verbal + visual info need to be integrated and stored somwhere in WM
• neuroimaging studies

Answer 11

• Baddeley et al., (1984) –> articulatory suppression decreases memory span for visual information but doesn’t eliminate it
  
  • lack of rehearsal should stop storage completely - yet can recall 5
  • must be temporarily stored somewhere (EB)
  • can’t just be using articulatory control processes to store visual info
• Chincotta et al., (1999) –> use combined verbal + visual representation when performing tasks with Arabic numerals but not with digit words
  
  • Arabic - used verbal + visual –> in integrated way
• Baddeley et al., (1987) - memory span for meaningful sentences = 15-16 words
  
  • more than PL capacity
  • draw on semantic memory to boost capacity?

BUT:

• Baddeley & Wilson (2002) - amnesiacs with impaired LTM have normal immediate sentence span –> not drawing on semantic memory
• Baddeley (2000) - recalls densely amnesic patients able to play bridge - requires memory

Neuroimaging studies:

• Prabhakaran et al., (2000) - fMRI of WM tasks –> just verbal; just spatial; or integration; recall
  
  • posterior regions (medial temporal gyrus) - material specific WM effects
    
    • consistent with PL + VS
  • right frontal cortex (medial/superior temporal gyrus) - retention of integrated information
    
    • need episodic buffer

Question 12

Central executive (WM)

• what is it?
• functions
• evidence
• Dysexecutive syndrome

Answer 12

CE - coordination, attention, problem solving –> coordinates cognitive processes

Functions (Smith & Jonides, 1999):

• switching attention betweentasks
• planning subtasks to achieve goals
• selective attention of certain stimuli + ignoring others
• monitoring + evaluating contents of other WM stores

Evidence: Baddeley (1996)

• hold 1-8 digits in mind while generating random series of key presses
• randomness decreases as digit span increases
• due to greater demands on CE - switching attention

Dysexecutive syndrome –> CE breaks down after frontal lobe damage

• Rylander (1939) - review:
  
  • disturbed attention, hard to grasp new task, personality change, increased distractability
• D’Esposito et al., (1995)
  
  • dorsolateral regions (in frontal lobes) important for EF
  • greater activity in dual-task than single-task conditions
• Duncan et al., (2000)
  
  • lateral frontal cortex - neural basis of general intelligence (g)

Question 13

Long-term memory:

• capacity

Answer 13

• Finite number of neurons + synapses
• humans can take in 100 stimuli per minute (~3 billion over 70yr lifetime)
• Standing (1973) - couldn’t find LTM limit - even for 10,000+ pictures - perform above chance
• Lindauer (1986) - conceptual approach
  
  • capacity contrained by rate of acquisition
• Ebbinghaus (1885) - retention decreases with longer intervals
  
  • rapid loss then decreases until stabilising point
  • see forgetting curve

Question 14

Long-term memory: maximising retention

• practice
• level of processing
• elaboration
• organisation
• spacing

Answer 14

Practice:

• Pirolli & Anderson (1985) - LTM retrieval improved with practice
  
  • linear growth of long-term potentiation with practice (fMRI)
  • as practice increased, recognition time decreased

Level of processing:

• Craik & Lockhart (1972) - processing affects memorability
  
  • greater accuracy if think of meaning
  • less accurate if just repetition
• Craik & Tulving (1975) - perceptual (captial letters?); phonological (rhymes); semantic (meaning)
  
  • semantic qualities helped accuracy of recall the most
  • deep-level processing boosts long-term retention

Elaboration - relate new material to associated info + prev. knowledge

• Anderson & Bower (1972) - elaboration condition improved memory from 57%-72%
  
  • because it was self-generated –> knowledge/details/explanations/e.gs/information

Organisation

• Bower et al., (1969) - group 1 = words in organisational hierachy; group 2 = randomly presented
  
  • recall almost doubled in 1
  • order of reproducing words mirrored the hierarchical representation (in memory)

Spacing - between coding episodes

• Massed practice =many trials with same stimulus undertaken without interruption
• spaced practice = increased intervals of time used between repetitions of stimuli
• Bahrick (1979):
  
  • massed study - better for immediate retention (STM)
  • spaced study - better for later retention (LTM)

Question 15

What causes forgetting (LTM) –> decay or interference?

Answer 15

• Jenkins & Dallenback (1924): sleep condition (decay) vs wake condition (decay + interference)
  
  • greater forgetting during wake than sleep
  • BUT: now we know sleep = memory consolidation
• Baddeley & Hitch (1977) - rugby players - some played 6 games, some only 1 (over 2 months)
  
  • if decay - memory of game should be same
  • if interference - 6 games should be worse at recalling
  • there was a correlation between recall + number of games –> interference

Question 16

Types of interference (LTM)

Answer 16

Proactive interference –> previous learning disrupts later learning

• Underwood (1957) - more previous lists learned - greater forgetting occurred

Retroactive interference –> later learning disrupts earlier learning

• Slamecka (1960) - learn sentences 2,4, or 8 times THEN either 0,4,or 8 different trials
  
  • amount learned was a function of initial learning (8>4>2)
  • forgetting was a function of number of sentences learned after (0>4>8)
• misinformation effect - distortion of memory by inserting other information (e.g. leading questions)

Question 17

Forgetting –> amnesia

• HM
• types
• tests

Answer 17

HM:

• medial temporal lobes removed (epilepsy), anterograde + retrograde amnesia (of ~3 years pre-surgery)
• no change in intellectual functioning, language comprehension, semantic knowledge or perception
• impaired explicit memory (demonstrative) –> impaired primacy effect
• intact implicit memory –> STM, recency effect, can learn new motor skills

Retrograde amnesia = memory loss of event prior to brain damage/surgery/lesion

Anterograde amnesia = can’t form new memories or retain new information

Tests:

• Rempel-Clower et al., (1996): amnesia patient impaired at recalling complex figure
  
  • Rey Complex Figure task
  • copied it fine (no motor/perceptual difficulty) BUT couldn’t recall later
• Hassabis et al., (2007): amnesia patients impaired at imagining the future too
  
  • patients with medial-temporal lobe amnesia/damage
  • impaired prediction/estimation of future

Question 18

Long-term memory: Squire (1992) diagram

Answer 18

• LTM –> explicit (declarative) + implicit (non-declarative)
• Explicit (conscious) –> episodic (personal events) + semantics (meaning/facts/knowledge)
• Implicit (unconscious) –> perceptual representation (priming) + procedural memory (motor skills/conditioning)

Question 19

Implicit vs explicit memory tasks (LTM)

• 2 potential issues

Answer 19

Explicit memory tasks:

• recognition (old/new)
• cued recall (all of category x)
• free recall (all from list)

Implicit memory tasks:

• fragment completion (part of word missing: a*pl*)
• word stem completion (complete word: ap-)
• degraded word naming (word partially obscured/masked)

Process purity –> important to note that even in explicit responses there may be implicit influences

Fallacy of reification –> treating a task as if it definitely taps into an actual system

Question 20

Explicit and implicit memory - 2 separate systems

• behavioural evidence
• neuropsychological evidence
• functioning neuroimaging studies

Answer 20

Behavioural:

• Jacoby & Dallas (1981) - words encoded using perceptual; phonological; or semantic processes
  
  • implicit memory shows little impact
  • explicit shows classic memory effect (see Craik & Tulving, 1975)
    
    • better accuracy for semantic
• Jacoby & Dallas (1981) - modality difference between study + test
  
  • explicit shows little impact
  • implicit shows significant reduction in accuracy
  • different modality - perceptual representation is different (so no priming)

Neuropsychological (medial temporal lobe amnesia):

• Warrington & Weiskrantz (1970) - pursuit rotor and Gollin figures task - performance increases over trials
  
  • preserved implicit memory even if no explicit memory of having done it
• Graf et al., (1984)
  
  • patients with amnesia impaired on recall + recognition (explicit) but not word completion
• Gabrieli et al., (1995) - MS (lesion affecting occipital cortex) vs amnesiacs
  
  • MS performed as well as controls on explicit memory tasks
  • MS impaired on implicit memory tasks
  • double dissociation from amnesiacs

Functional neuroimaging:

• Schacter & Wagnr (1999) - medial temporal lobe activity in explicit encoding and retrieval
• Simons et al., (2003) - priming associated with activation of bilateral fusiform cortex
  
  • activity along visual cortex = ventral visual stream
  • early in visual stream –> priming is very perceptual
  • as activity moves along temporal lobe base - priming is more semantic

Question 21

Episodic and semantic memory (explicit LTM)

• what are they
• evidence for distinction
  
  • behavioural
  • neuropsychological

Answer 21

Episodic memory - about specific events in specific time + place; specific to individual

Semantic memory - general knowledge about objects/people/fact/concepts + word meanings

Behavioural evidence:

• Jacoby & Dallas (1981) - processing at encoding affects episodic but not semantic memory
  
  • accuracy the same for semantic memory if encoded based on perception, phonology, semantics
• Jacoby (1983) - at encoding, reading word out of context improved seemantic memory
  
  • generating word themselves improved episodic memory

Neuropsychological evidence:

• HM - episodic impaired, semantic not
• Vargha-Khadem et al., (1997) - children with selective hippocampal damage
  
  • impaired episodic memory, good semantic memory (vocab)

Question 22

Tulving’s SPI model:

• what it is
• ISSUES - Dementia

Answer 22

serial encoding (s); parallel storage (p); independent retrieval (I)

• can have semantic withough episodic
• can’t have episodic without semantic
• perceptual representation semantic memory episodic memory

BUT: Dementia

• Early Alzheimer’s - medial temporal lobe atrophy
  
  • impaired episodic, intact semantic
• Semantic dementia - lateral temporal lobe atrophy
  
  • degradation of semantic knowledge
• Simons et al., (2002) - double dissociation between AD + SD:
  
  • AD - no episodic, ok semantic
  • SD - ok episodic, no semantic

So…Multiple Input Hypothesis

Question 23

Adapting Tulving’s SPI model –> Multiple input hypothesis

Answer 23

Graham et al., (2000) - perceptual feeeds episodic memory too

• controls - rely on perceptual + semantic controls to drive episodic memory
  
  • if one is missing, can draw on the other
• SD - did just as well as controls in perceptually identical condition
  
  • only impaired when perceptually different - because can’t rely on perceptual representation