the multi-store model of memory (Atkinson and Shiffrin, 1968) Flashcards
1
Q
what does the msm consist of?
A
- three basic stores:
- sensory register
- stm
- ltm
2
Q
what is a sensory experience?
A
- something we’ve seen, heard, touched, etc first enters the sensory register/memory where it’s held for brief moment before decay
3
Q
where does the attended information from the sensory register go?
A
- it’s transferred to the short term store
4
Q
how long is the info held for in the stm?
A
- for around 30 seconds before decay
- unless rehearsal is used to maintain this info for a longer period of time
5
Q
where can info from stm be transferred to?
A
- long term permanent storage in the long term store
6
Q
sensory register/memory
A
- we experience the world through our senses
- visual sensory register has been researched using a whole or partial report technique (Sperling) in which a visual array of letters is presented via a tachistoscope for a brief moment
- then a direction is given to recall whole or specific row of array
- ppts recalled on average 4.32 letters of whole array
- shows that sensory register can only hold limited amount of info for only a few hundred milliseconds before it’s lost
7
Q
what is a tachistoscope?
A
- a device used to present info in a controlled way
8
Q
what is the duration of stm?
A
- info that’s been attended enters the stm and held temporarily for 15-30 seconds and then decays unless it’s maintained through rehearsal
9
Q
what did Peterson and Peterson investigate? (duration of stm)
A
- investigated duration of sym using trigram interference task to prevent rehearsal
- ppts were required to remember a single trigram of three consonants for intervals of 3,6,9,12,15 and 18 seconds
- the trigram was read out and ppts then were given a number from which they had to count backwards in threes eg 679, 676, 673
- correct recall of trigram was likely after short interval but performance dropped rapidly after 15-18 secs
- concluded that decay occurs in stm over period of 15 secs
10
Q
what is the capacity of stm?
A
- capacity = amount of info that’s can be stored by the memory system
- Miloer showed that stm is limited to around seven items of info
- ’Magic number seven, plus or minus two’
- can view the stm as a series of between five to nine slots in which info can be stored
11
Q
what is the primary-recency effect?
A
- Glazer and Cunitz
- found support for msm in study
- using word lists they found the first words in list were recalled well, as were the last words
- but middle words weren’t remembered well
12
Q
what do they say about the primary effect?
A
- info learned first (primary effect) is well remembered because it’s gone into LTM through the rehearsal loop
13
Q
what do they say about the recency effect?
A
- info learned at the end (recency effect) is well remembered because it’s still in rehearsal loop
- therefore available for immediate recall but info in the middle is not recalled recalled because it didn’t go from rehearsal loops into ltm as it was displaced by new material coming into loop
- therefore info was lost - forgotten
14
Q
what is encoding in stm?
A
- encoding = transforming sensory experience into a form that can be held/used by memory system
- Atkinson and Shiffrin believed that memory trace in stm was held in a auditory or verbal form because of phonological similarity effect
- phonological similarity effect = letters and words of a similar sound presented to ppts are more difficult to recall than dissimilar sounds letters and words
- similarity of sounds leads to confusion in STM suggesting encoding in this store is primarily acoustic: auditory or verbal
15
Q
what is the retrieval from stm?
A
- based on rapid scanning of stored info
- rehearsal is important in maintaining info in stm, increasing strength of memory trace and ultimately building up memory trace in ltm
- digit span exps suggest we’re able to maintain between 5 and 9 items using rehearsal
- as more info is input into store, other info is knocked out (displaced and quickly decays
16
Q
transfer of info between stm and ltm
A
- in order to transfer info received by sensory register to stm, we must use our ltm to make sense of the info and assign it a verbal label
- eg we may register image of horse, but this can’t be stored as auditory-verbal stm until we’ve identified it using ltm of what object represents
- transfer of info from st to lt store can be as result of rehearsal, although this would leave relatively weak memory trace
- stronger memory trace can be achieved by using mental operation, eg mnemonic to increase strength of transfer
17
Q
retrieval from ltm
A
- Atkinson and Shiffrin believed that Tom’s exist for all sensory modalities
- we have memories for taste, sound, smells etc
- they said that multiple copies of a memory were retained in ltm, not just a single memory trace
- Tip-of-the-tongue phenomenon (Brown and McNeill) supports idea as they showed that people were able to accurately predict that they could recognise correct answer even if they couldn’t recall answer at that moment in time
18
Q
what is the encoding in ltm?
A
- can depend on rehearsal process or some form of association between new and pre existing knowledge
- if info is linked to pre existing knowledge it makes search for info far easier
- makes sense as a random search of such large store would be exhaustive
- encoding in ltm is semantic - encode through meaning
19
Q
what is the duration of ltm?
A
- potentially a lifetime
- Bahrick et al investigated what they referred to as Very Long-term Memory (VLTM)
- used series of memory tests on names and faces of students in hugh school yearbooks
- 400 ppts between ages 17-74 were used
- found that identification of names and faces was 90% accurate within 15 years of leaving school and between 70-80% accurate 48 years after leaving school
- shows that although memory deteriorates over time, ltm for faces and names is fairly resilient over passage of time
20
Q
what is capacity of ltm?
A
- potentially infinite
- Brady showed lots 2500 objects over course of 5.5 hours
- later when pots were shown original object paired with very different object, identification was 92% and if different object was similar identification was 88%
- demonstrated that thousands of images can be maintained successfully in ltm