The Multi-Store Model (MSM) of Memory

Question 1

what are the three stores in the memory model

Answer 1

The model consists of three SEPARATE unitary stores: the short-term store, long-term store and the third store in this model is known as the sensory register. The sensory register contains information collected by your senses, for example information received from the ear, eyes and nose.

Question 2

what kind of approach does the multi-store model take

Answer 2

The model takes an information processing approach to explaining the memory system, whereby the system is characterised by a set of stages and information flows through each stage in a fixed linear sequence.

Question 3

what are the two limitations at each stage

Answer 3

There are capacity and duration limitations at each stage.

Question 4

sensory store

Answer 4

The sensory stores constantly receive information but most of it receives no attention. This information remains in the sensory store for a very brief period before decaying (fading away) or if given attention then the data is transferred to the short-term memory.

Question 5

Memory in STM

Answer 5

Memory traces in STM are fragile and unless rehearsed (maintenance rehearsal) they can be lost within about 30 seconds through displacement (pushed out by incoming information) or decay (memory trace fades completely). Information is typically coded acoustically.

Elaborative rehearsal enables the transfer of information between STM and LTM.

Question 6

Memory in LTM

Answer 6

Memory traces in the LTM can be retained for a lifetime, although may be subject to loss through decay (memory trace fades), retrieval failure (material is available but not accessible, a cue may be needed to trigger the retrieval of the memory) or interference (confusion with other memory traces).

Question 7

coding in sensory register

Answer 7

Coding in the Sensory Register
The SR processes sensory information received from the sense organs e.g. eyes, ears, nose.
Information is stored in a raw, unprocessed form, with separate sensory stores for different sensory inputs (modality specific):

Echoic store – for auditory information
Iconic store – for visual information
Haptic store – for tactile (touch) information
Olfactory store – for smells

Question 8

coding in SR

Answer 8

Coding in the SR is modality specific.
Crowder (1993) found that the SR only retains information in the iconic store for a few milliseconds, but for 2-3 seconds within the echoic store.
This suggests that sensory information is coded into different sensory stores (modality specific).
Furthermore it suggests that they have different durations.

Question 9

capacity of the SR

Answer 9

The capacity of each store in the SR is very large, with information in a highly detailed form.
The fact that the SR has such a short duration makes it difficult to research its capacity as the information leaves the store so quickly.

Question 10

Duration of the SR

Answer 10

All sensory memory stores have limited duration (250-500 milliseconds), although the duration varies from store to store, with different types of information decaying at different rates.
E.g. Walsh and Thompson (1978) found that the iconic sensory store has an average duration of 500 milliseconds, which decreases as individuals get older.
Evolutionary advantage of the brief duration of the SR: People can focus on perceptual information, with an immediate survival value. Retaining non-useful information would compromise this.

Question 11

The Sensory Register – Research Evidence
Sperling (1960)
Aim:

Answer 11

To investigate the capacity of iconic memory

Question 12

The Sensory Register – Research Evidence
Sperling (1960)
Method:

Answer 12

A three by four grid of numbers was flashed onto a screen for 0.05 seconds, followed by a high, medium or low pitched tone to indicate which row was to be recalled

Question 13

The Sensory Register – Research Evidence
Sperling (1960)
Results:

Answer 13

On average, the participants were able to recall 80% of the letters on the cued row

Question 14

The Sensory Register – Research Evidence
Sperling (1960)
Conclusions:

Answer 14

Since the participants didn’t know which row was going to be called beforehand but still managed to recall it well, you can assume that at one time all of the information was held in the sensory memory, although it decayed very rapidly.

Question 15

The Sensory Register – Research Evidence
Sperling (1960)
Evaluation

Answer 15

The work by Sperling led to the view that sensory memory stores are large but decay very rapidly, lasting 250 to 500 milliseconds.

Question 16

Overview: Short-Term Memory
Capacity

Answer 16

Limited
Miller argued that most people can store 5-9 items in STM.
Digit span = 7+/-2 items. Miller’s ‘magic number 7’.
Tested by Jacobs (1987) using the serial digit span technique - see next slides for detail about procedure and findings.
Capacity can be enhanced through ’chunking’ – this is where individual pieces of information (chunks) are grouped into larger units, therefore taking up less space in the STM.

Question 17

Overview: Short-Term Memory
Duration

Answer 17

Limited
18-30 seconds
Peterson and Peterson (1959) – ppts only recalled about 2% of trigrams when there was an 18 second time interval, compared to about 90% after a 3 second interval – see next slides for more detail.

Question 18

Overview: Short-Term Memory
Coding

Answer 18

Mainly acoustic (although recent research has suggested that information is also encoded visually in STM)

Question 19

Baddeley (1966): Coding in STM and LTM
Aim

Answer 19

To assess whether coding in STM and LTM is mainly acoustic (by sound) or semantic (by meaning).

Question 20

Baddeley (1966): Coding in STM and LTM
Procedure

Answer 20

75 ppts were divided into 4 groups and were presented with one of four word lists (consisting of 10 words), repeated four times:
List A: Acoustically similar words e.g. ‘cat’, ‘mat’, ‘sat’
List B: Acoustically dissimilar words e.g. ‘pit’, ‘day’, ‘cow’
List C: Semantically similar words e.g. ‘big’, ‘huge’, ‘tall’
List D: Semantically dissimilar words e.g. ‘hot’, ‘safe’, ‘foul’
STM: Participants were then given a list containing the original words in the wrong order. Their task was to rearrange the words in the correct order.
LTM: the procedure was the same, but there was a 20 minute interval before recall, during which ppts performed a distractor task to prevent rehearsal.

Question 21

Baddeley (1966): Coding in STM and LTM
Findings

Answer 21

STM: Ppts given List A (acoustically similar) performed the worst, with a recall of only 10%. They confused similar sounding words e.g. ‘cap’ instead of ‘cat’. Recall for the other lists was comparatively good at between 60-80%.
LTM: Ppts given List C (semantically similar) performed the worst, with a recall of only 55%. They confused similar meaning words e.g. ‘big’ instead of ‘huge’. Recall for the other lists was between 70-85%.

Question 22

Baddeley (1966): Coding in STM and LTM
Conclusiuons

Answer 22

STM: Baddeley concluded that STM is coded on an acoustic basis, due to the acoustic confusion.
LTM: Baddeley concluded that LTM is coded on a semantic basis, due to the semantic confusion

Question 23

Baddeley (1966): Coding in STM and LTM
Evaluation

Answer 23

:)This was a laboratory study and therefore shows causality - cause and effect can be established.
:(However, this study may lack ecological validity because the tasks are not representative of everyday activities.

Question 24

Brandimonte et al. (1992) found

Answer 24

Brandimonte et al. (1992) found evidence to support this.
They presented ppts with six line drawing of familiar objects and asked them to memorise each one.
It is thought that the drawings were encoded acoustically in terms of the object’s name, so when ppts were asked to form a mental image of the drawing and subtract a part of it to reveal a different object, they were less able to name the image than those ppts who were asked to memorise the drawings whilst repeating the meaningless chant ‘la-la-la’ (which would have suppressed acoustic encoding, thus encouraging visual encoding of the initial drawing instead

Question 25

what did Baddeley’s research suggest

Answer 25

Baddeley’s research suggests that coding in STM is acoustic, however other research has suggested information can also be encoded visually in STM.

Question 26

Jacobs (1887): Capacity in STM
Aim

Answer 26

To assess the capacity of short-term memory

Question 27

Jacobs (1887): Capacity in STM
Procedure

Answer 27

Laboratory experiment
Recall was tested as follows:
The experimenter used serial digit span technique
Increasingly long lists of numbers and letters were read out to participants
After each list was read out, participants were required to recall the list in the order they heard the numbers or letters e.g. ‘8, 5, 3, 9’ (followed by recall), ‘2, 4, 7, 1, 3’ (followed by recall)
When participants failed on 50% of tasks, they were judged to have reached their capacity

Question 28

Jacobs (1887): Capacity in STM
Findings

Answer 28

Jacobs found a difference between capacity for numbers and letters:
Jacobs found capacity for numbers was 9 items
He found capacity for letters was 7 items
He also noticed that recall seemed to increase with age - eight year olds being able to recall an average of 7 digits whereas by the age of 19 recall had increased to 9 digits.

Question 29

Jacobs (1887): Capacity in STM
Conclusions

Answer 29

STM has a limited capacity of between 5 and 9 (7 +/-2) items of information.
As age increases, we appear to develop better strategies of recall.

Question 30

Jacobs (1887): Capacity in STM
Evaluation

Answer 30

:)Later studies (Miller, 1956) have supported Jacobs’ findings and conclusions: “The magic number seven, +/- 2.”

:)Miller and others have also discovered that chunking can increase capacity (BBC or 1066 becomes one chunk of information rather than 3 or 4 distinct chunks).

:(Lacks ecological validity: Testing the capacity of STM by recalling lists of numbers/letters bears little relevance to everyday activities involving the STM; the task therefore lacks mundane realism.

:(Furthermore, there is no consistency across different stimulus material.It seems that with numbers we can indeed typically recall seven.However, this drops to six for letters and is nearer five for words.

Question 31

Peterson and Peterson (1959): Duration in STM
Aim

Answer 31

To assess the duration of short-term memory, when verbal rehearsal is prevented

Question 32

Peterson and Peterson (1959): Duration in STM
Procedure

Answer 32

Laboratory experiment; 24 university students were tested
Recall was tested as follows:
The experimenter said a nonsense trigram to the participant e.g. WRT, and then said a three-digit number
The participant had to count backwards from this number in 3s or 4s until told to stop – this was to prevent rehearsal of the trigrams (rehearsal would have encouraged the information to remain in STM for longer)
The participant was then asked to recall the nonsense trigram (in the order they heard the letters)
Each participant was given 2 practice trials followed by 8 trials
On each trial, the retention interval (time spent counting backwards) was different: 3, 6, 9, 12, 15 or 18 seconds

Question 33

Peterson and Peterson (1959): Duration in STM
Findings

Answer 33

Participants remembered more trigrams accurately when they only had a 3-second interval and recalled fewer trigrams when the retention interval was 18 seconds.
They remembered about 90% of the trigrams when there was only a 3-second interval and about 2% when there was an 18-second interval.

Question 34

Peterson and Peterson (1959): Duration in STM
Conclusions

Answer 34

Information remains in STM for less than 18 seconds if verbal rehearsal is prevented

Question 35

Peterson and Peterson (1959): Duration in STM
Evaluation

Answer 35

:(Lacks ecological validity: Testing the duration of STM using a string of meaningless consonants in a laboratory setting is not something you would do in everyday life; the task lacks mundane realism.

:(Lacks population validity: Only students participated in this research – it is likely that they have rather different memories compared to people of other age groups; students may be more than averagely intelligent

Question 36

Overview: Long Term Memory
Coding

Answer 36

mainly semantic (by meaning)
Baddeley (1966) – recall was worse for semantically similar words than semantically dissimilar words. Recall was the same for acoustically similar and acoustically dissimilar words

Question 37

Overview: Long Term Memory
Capacity

Answer 37

potentially unlimited
Research has suggested that LTM is divided into a number of different memory systems, although this was not recognised by the Multi-Store Model of memory.
Clive Wearing – demonstrated the existence of the procedural memory.

Question 38

Overview: Long Term Memory
Duration

Answer 38

potentially a lifetime
Bahrick et al. (1975)

Question 39

Case Studies – Clive Wearing

Answer 39

The case of Clive Wearing both supports and challenges the multi-store model of memory….
:)Clive Wearing was unable to lay down any new long term memories, after his hippocampus was destroyed by a viral infection, suggesting that there is a separate short-term and long-term memory store in the brain.

:(However, Clive was still able to sight-read and play complex pieces on the piano, suggesting that his procedural memory was intact, which cannot be explained by the multi-store model of memory as it doesn’t recognise different types of long-term stores.

Question 40

Bahrick (1975): Duration in LTM
Aim

Answer 40

To establish the existence of very long-term memory and to see whether there was any difference between recognition and recall.

Question 41

Bahrick (1975): Duration in LTM
Procedure

Answer 41

Investigators tracked down the graduates from a high school in America over a 50-year period.
392 graduates were shown photographs from their high-school yearbook.
Recognition group: for each photo participants were given a group of names and asked to select the name that matched the photo.
Recall group: participants were simply asked to name the people in the photos without being given a list of possible names.

Question 42

Bahrick (1975): Duration in LTM
Findings

Answer 42

In the name-matching (recognition) condition, participants were:
90% correct 14 years after graduation
80% accurate after 25 years
75% accurate after 34 years; 60% accurate after 47 years
In the recall group, participants were not as successful:
60% accurate after 7 years
Less than 20% accurate after 47 years

Question 43

Bahrick (1975): Duration in LTM
Conclusions

Answer 43

People can remember certain types of information for a lifetime.
Long-term memory appears to be better when measured by recognition tests than by recall tests

Question 44

Bahrick (1975): Duration in LTM
Evaluation

Answer 44

:)High ecological validity: This study used meaningful stimulus material (high-school yearbooks) and tested people for memories from their own lives.
:(It is unclear whether the drop-off in accuracy after 47 years reflects the limits of duration or a more general decline in memory with age.

Question 45

Research evidence to support the idea of a separate STM and LTM

Answer 45

The Serial Position Effect in a free recall task
Glanzer and Cunitz (1966) – On immediate recall ppts tended to recall the words from the beginning of a word list (primacy effect) and from the end of the list (recency effect), rather than in the middle.However, following a time delay the recency effect disappears.
Primacy effect – words at the beginning are rehearsed and transferred to LTM.

Recency effect – these words remain in the STM prior to recall.

(Words in the middle of the list are displaced by words towards the end of the list)

Question 46

Areas of the brain associated with STM and LTM

Answer 46

Research has found that the prefrontal cortex is active when individuals are working on a task in immediate (i.e. short-term) memory (Beardsley, 1997).
However, the hippocampus is active when long-term memory is engaged (Squire et al., 1992).

Question 47

What does this research suggest and how does it add support to the multi-store model of memory?

Answer 47

It adds support to the idea that STM and LTM are separate stores in that it suggests that they are located in different areas of the brain.

Question 48

Multi-Store Model: Case Studies – H.M

Answer 48

Similar to Clive Wearing, the case of H.M. both supports and challenges the multi-store model of memory….

:) The fact that H.M. was unable to form any new long term memories, after the removal of his hippocampus, suggests that there’s a separate short-term and long-term store.

:( However, H.M was able to acquire new skills, suggesting that his procedural memory was intact. This cannot be explained by the multi-store model of memory as Atkinson and Shiffrin viewed the LTM as a unitary store.

Question 49

:)The multi-store model gives a reasonable account of the structure of memory
:)P:Research support – the following studies all support the idea that there is a separate STM and LTM

Answer 49

• Case studies: H.M; Clive Wearing
• Laboratory study: Glanzer and Cunitz (1966)
  E: Glanzer and Cunitz: Ppts were found to show better recall for the words from the beginning of a word list (primacy effect) and from the end of the list (recency effect), rather than in the middle.
  The primacy effect provides evidence for the existence of the LTM i.e. information has been rehearsed and transferred into the LTM.
  The recency effect provides evidence for the existence of the STM i.e. the information is recalled immediately, so is still present in the STM.
  C: The effects observed support the idea of a separate STM and LTM.

Question 50

Evaluating the Multi-Store Model – Limitations

Answer 50

It doesn’t really explain the processes involved in coding - there is too much emphasis on the structures in the memory system.

STM relies on LTM – this is not recognised by the model as it views the STM and LTM as independent stores.
Ruchkin et al. (2003) – STM is the part of the LTM that is activated when necessary.

Rehearsal may not be the only means by which to commit memories to the long term memory.
It might be the case that information is transferred to the LTM because of its distinctiveness or significance to the person (Eysenck, 1979).
For example, flashbulb memories are memories for the circumstances in which one first learned of a very surprising and consequential (or emotionally arousing) event.

Question 51

:(The model has been criticised for being too reductionist and inflexible to explain the entire memory system.

Answer 51

It is reductionist because it reduces the complex nature of memory down to a simple set of ideas.
It is an oversimplification of the memory structures and processes.
For example, it doesn’t really explain the processes involved in coding – there is too much emphasis on the structures e.g. SR, STM, LTM.

The Working Memory Model (WMM) arguably provides a more comprehensive and detailed model of memory, challenging the view that STM is a single (unitary) store, instead suggesting it is an active processor

Question 52

Multi-Store Model: Additional Limitations

Answer 52

 The case study of K.F. challenges the STM as a single (unitary) store, as does the Working Memory Model.
Evidence (H.M., Clive Wearing) has also suggested that there are different types of long term memory e.g. declarative, episodic, procedural, and that this isn’t a single store as the multi-store model would suggest.
For example, Clive Wearing’s procedural memory was intact, however, his episodic memory was affected by the damage to his hippocampus.