Architecture of human memory

Question 1

Sensory stores

Answer 1

Large capacity, v short duration, allows quick/ online interaction with the environment
All incoming info is held v briefly (1/2-2 seconds) in as a copy of the actual sensory information.
Information is primitive & unanalysed. Evidence suggests precategorical.

e.g ‘Iconic’ memory and ‘echoic’ memory aka precategorical acoustic store.

Precise lengths of different sensory memories differ e.g visual < auditory.

Considerable evidence suggests it is the info we pay attention to that moves on to further processing.

Question 2

Iconic memory

Partial report?

Answer 2

Holds visual info.

Sperling 1960: Partial report procedure showed that iconic memory holds 9-12 items, but difficult to measure capacity because icons are fading as attempt to measure
Partial report advantage:
- only with sensory cues (colour, location, tone), not if cued by category (letters, numbers): suggests store is precategorical
- temporary & unstable- works only if interval is short, and dark & blank (not masked by new, distracting stimuli)

Question 3

Echoic memory

Suffix effect

Answer 3

Aka ‘Precategorical Acoustic Store’: PAS

Neath et al 1993:

• Stimulus suffix effect (e.g digit 0 spoken by examiner) dramatically reduces recall of the final item.
• PAS explains stimulus suffix effect by assuming that suffix enters into PAS & interferes with final item
• visual suffix will generally have little/ no effect on auditorily presented list of items.
• context dependent suffix effect: larger suffix effects obtained when suffix labelled as speech rather than nonspeech, despite suffix being same physical token in both cases.

Crowder & Morton 1969:

• newly entering items interfere with already present items only if acoustically similar. Size of suffix effect presumed to be based solely on acoustic similarity
• PAS is precategorical- top down knowledge/ categorisation of suffix theoretically cannot alter magnitude of decrement

Question 4

Short term store:
Duration?
Capacity?
Fragility of storage?

Answer 4

15-30s
Capacity = 7 items/ ‘chunks’, plus or minus 2. e.g Miller 1956, v consistent finding
Rehearsal with distraction prevents retention (Peterson & Peterson)

Question 5

Chunk

Answer 5

= a meaningful unit of info

Question 6

Digit span

Answer 6

about what one can say in 1.8 seconds

Digit span for Chinese > English > Welsh

Question 7

Similarity effects on STM

Answer 7

Baddely et al 1966:
Subjects perform worse in intermediate recognition tasks with phonologically confusing lists than on non-confusable lists.

Question 8

Working memory model (replaced STM)

Answer 8

Baddeley & Hitch 1974, Baddeley 1986, 2000

Multi-component, ‘active’ working memory system.

Central executive connected reciprocally to Phonological loop, visuo-spatial sketchpad, episodic buffer (‘slave’ system, added 2000)

All components have limited capacity & are independent of other components

Question 9

Phonological loop

Answer 9

aka articulatory loop

Holds info in speech based form
Preserves order in which words are presented

2 parts:

• Short term phonological store with auditory memory traces subject to rapid delay. Info coded & stored for ~2 secs
• Articulatory rehearsal component that can revive the memory traces. Stored info refreshed by subvocal rehearsal.

Evidence:

• Phonological similarity effect (Larsen et al 2000)
• Word length effect (Baddeley et al 1975)
• Articulatory suppression (Baddeley 1984)
• Unattended speech effect (Colle & Welsh 1976, Salame & Baddeley 1987)

Question 10

Visuo-spatial sketchpad

Answer 10

Used for temporary storage & manipulation of spatial & visual info
Specialised for spatial & visual coding

Evidence: Baddeley & collegues investigated using Brooks matrix task & dual task paradigm. Found that

• without concurrently performing a visuo-spatial tracking task, subjects performed equally on nonsense & spatial instructions.
• when performing tracking task, performance with spatial instructions was extremely poor

Question 11

‘Dual task’ paradigm

Answer 11

Assumptions:

• if two tasks use the same component, they cannot be performed successfully together
• if two tasks use different components, it should be possible to perform them as well together as separately.

Baddeley & Hitch: investigated the involvement of 3 components of working memory (central exec, sketchpad & loop) in selection of chess moves by weaker & stronger players.
Selection of chess moves involved central exec & vs sketchpad but not phonological loop. Similar effects in strong & weak players.

Question 12

Brooks matrix task

Answer 12

Subjects task = to repeat back a series of instructions.
In one condition, statement can be recorded into a path through the matrix, and therefore remembered as a specific pattern.
In second condition, adjectives presented are not suitable for spatial encoding

Brooks demonstrated that spatial sequences are best retained when presented auditorily, nonsense sequences best retained when presented visually.

Question 13

Central executive

Answer 13

Most complex, least understood component of WM
Coordinated activity of 2 (/ 3 with episodic buffer) slave systems

Other major functions:

• switching of retrieval plans
• timesharing in dual task studies
• selective attention to certain stimuli while ignoring others
• temporary activation of LTM

Evidence: Baddeley 1996: damage to frontal lobes causes impairment of CE. Patients suffer from dysexecutive syndrome- behave as if lack a control system enabling them to direct & re-direct their processing resources appropriately

Question 14

Double dissociation

Answer 14

Involves 2 patients: A and B, and 2 tasks: I and II
Occurs when patient A performs normally on task I but is impaired on task II, and patient B performs normally on task II but is impaired on task I.
Allows us to infer that tasks I and II involve separate cognitive functions. If A and B exhibit damage in different parts of brain, offers evidence as to where those cognitive functions are located

Question 15

Double dissociation of LTM and STM

Answer 15

HM: medial temporal damage. Impaired LTM, normal digit span
KF: damage in left parieto-occipital region. No problem with long term learning and recall. Digit span greatly impaired, had recency effect of only 1 item.

Question 16

Recency effect

Answer 16

Last few items in a list often much better remembered than items from middle of list
Attributed to last few items being in short term store.
Can be eliminated if if participants counted backwards prior to recall (Glanzer & Cunitz 1966)

Question 17

Primacy effect

Answer 17

First few items in list remembered better than those in middle.
Not affected by interference after list finished
Attributed to LTM
Affected by word familiarity, presentation rate, age of participant.

Question 18

Long term memory
Capacity & duration

Proactive vs retroactive interference

Answer 18

In theory, unlimited capacity, but finite number of neurones & synaptic connections suggests some limit.
Lindaur (1986) suggested capacity constrained by rate of acquisition. Humans can take in max ~100 stimuli per min- thus theoretical limit of 3bn stimuli over 70 year lifetime.

Forgetting v rapid over first 1hr after acquisition, then reduces approx logarithmically
Evidence suggests forgetting primarily due to interference
- Proactive interference: previous learning interferes with later learning
- Retroactive interference: later learning disrupts earlier learning.

Question 19

Long term memory

Division into separate systems

Answer 19

Explicit (conscious)
- Episodic (personal events)
- Semantic (facts, knowledge)
Implicit (not conscious)
- Perceptual representation system (priming)
- Procedural memory (motor skills, conditioning)

Question 20

Explicit vs implicit memory

Answer 20

Explicit: Requires conscious recollection of previous experience
Implicit: facilitated in the absence of conscious recollection

Jacoby & Dallas 1981: deeper level of processing improved explicit memory, little effect on implicit.

Question 21

Episodic vs semantic memory

Answer 21

Episodic= memory of specific events/ episodes occurring in a particular time & place. Specific to each person
Semantic = general knowledge without awareness of when/ where learned. Can be culturally shared. 

Jacoby & Dallas: Different levels of processing at encoding improved episodic memory, had no effect on semantic

Jacoby, a double dissociation:

• reading a word out of context at encoding improved semantic memory
• generating word themselves improved episodic memory

Question 22

Semantic dementia vs Alzheimers disease

Answer 22

Semantic dementia: Selective deficit in semantic memory. Difficulty remembering facts about a particular object e.g function & name

Alzheimers: selective deficit in episodic memory. Patients know what object is & what for, but don’t remember having seen one recently.