Multi Store Model of Memory

Memory

Process of recalling or reproducing information that has been previously learned & retained through associative mechanisms

Capacity

Amount of information that can be held in a memory store

STM = Limited

LTM = Infinite

Sensory Register = Very large

Duration

The length of time information can be held in memory

STM = Does not last very long

LTM = Potentially lasts forever

Sensory Register = Milliseconds (very brief)

Coding

Format in which information is stored in the various memory stores

STM = Acoustically

LMT = Semantically

Digit Span and Capacity of STM (Jacob)

• Attention test to see how many number a person is able to remember

• Jacob’s digit span test (1887)

  • Easier to recall digits as there are only 9, whereas there are 26 letters

Mean span for digits & letters

Mean span for digits = 9.3 items

Mean span for letters = 7.3 items

Magic No. and Capacity of STM (Miller)

• Capacity of STM is about 7 items ( 7 + - 2)

• 5-9 items

• We could hold more items in STM by chunking

Chunking

Grouping sets of digits or letters into units or chunks

Limitation 1 of Capacity (More limited)

Point: More recent evidence to suggest STM is more limited than Miller assumed

Evidence: Cowan (2001) conducted similar experiment to Miller’s digit span test & found that participants on average remembered 4 items

Explain: Cowan’s findings challenge Miller’s notion of STM capacity as being around 7 items. Instead Cowan suggest STM may only hold approx. 4 chunks of info.

Link: Therefore, lower end of Miller’s range may be more appropriate, suggesting that the typical capacity of STM is closer to 5 items rather than 7.

Limitation 2 of Capacity (Individual Differences)

Point: Capacity of STM is not same for everyone, so lacks objectivity

Evidence: Jacobs also found that recall (digit span) increased steadily w. age ; 8 yr olds could remember on average 6.6 digits whereas the mean for 19 yr olds was 8.6 digits

Explain: This age increase might be due to changes in brain capacity, and/or to the development of strategies such as chunking

Link: This suggests that capacity of STM is not fixed & individual differences may play a role

Duration of STM study PETERSON & PETERSON (1959) - aim & method

Aim : To investigate the duration of STM

Method: Laboratory Experiments, 24 undergraduate particiapnats, 8 trials

Duration of STM PETERSON & PETERSON (1959) - Procedure

• Participants presented w. consonant syllables / nonsense trigrams such as HGX, PRL as well as 3 digits (e.g. 421)

• Were asked to recall these letters after an interval of 3, 6, 9, 12 or 15 seconds (retention interval)

• During this interval, participants were asked to count down from the no. they were given

  • Prevented maintenance rehearsal

Duration of STM PETERSON & PETERSON (1959) - Findings

• After 3 seconds - 90% of participants recalled trigrams correctly

• After 9 seconds - 20% of participants recalled trigrams correctly

• After 18 seconds - 2% of participants recalled trigrams correctly

Duration of STM PETERSON & PETERSON (1959) - Conclusion

• As duration increased, recall decreased

• Suggests STM only has a duration of 18-30 seconds before it disappears

Duration of LTM BAHRICK ET AL (1975) - Aim and Method

Aim : To investigate the duration of LTM

Method: Field experiment, 400 participants, aged 17-74

Duration of LTM BAHRICK ET AL (1975) - Procedure

• Participants tested on their memory of classmates from their year

• Recall was tested through 2 ways:

  • Cue recall: Recognize photos from the year book (50 photos)

  • Free recall: Recall as many as they could

Duration of LTM BAHRICK ET AL (1975) - FIndings

• Results for Photo Recognition:

  • Participants tested within 15 years of graduation, had 90% accurate recall

  • Participants tested after 48 years of graduation, has 70% accurate recall

• Results for Free Recall

  • Post 15 years = 60% recall

  • Post 48 years = 30% recall

Duration of LTM BAHRICK ET AL (1975) - Conclusion

Bahrick et al concluded that long term memory could potentially last forever

Limitation 1 of Duration of STM (meaningless)

Point: Weakness of Peterson & Peterson is inclusion of meaningless stimuli in their study

Evidence: Nonsense trigrams e.g. HGX, PRL

Explain: Trying to memorize consonant syllables doesn’t truly reflect most everyday memory activities where what we are trying to remember is meaningful

Link: This means that although the task was artificial, the study does have some relevance to everyday life

Limitation 2 of Duration of STM (results may be due to displacement)

Point: A criticism of Petersons’ study is that it didn’t measure what it was supposed to

Evidence: Reitman (1974) used auditory tones instead of numbers so that displacement wouldn’t occur (sounds don’t interfere w. verbal rehearsal) and found that the duration of STM was longer

Explain: In the Petersons’ study participants were counting the numbers in their STM and this may displace or ‘overwrite’ the syllables to be remembered.

Link: This suggests that forgetting in the Petersons’ study was due to displacement rather than decay.

How is information coded in the sensory register

Sense specific

Iconically (visual info is coded visually)

Echoic (sound / auditory info is coded acoustically)

Research on coding BADDELEY (1996) - Aim & Method

Aim: To test the effects of acoustic & semantic similarity on STM and LTM

Method: Used word lists like cat, cab, can, cad, cap, mad, max, mat, man, mat in order to test the effects of acoustic & semantic similarity on STM and LTM.

• Learning phase separated from testing phase by an inference test

• IV = acoustic, semantic setting

• DV = Score on a recall test of 10 words

Research on coding BADDELEY (1996) - Findings & Conclusion

Findings: Found participants had difficulty remembering acoustically similar words in STM but not LTM

• Semantically similar words posed little problems for STM but led to muddled LTMs.

Conclusion: Suggests STM is largely encoded acoustically whereas LTM is largely encoded semantically

Limitation 1 of Research on coding - Mundane realism

Point: One limitation in Baddeley’s research is the lack of mundane realism, particularly in the experimental tasks used to assess memory processes.

Evidence: In Baddeley’s experiment, when testing long term memory participants were given a relatively short break of 20 minutes between encoding & retrieval tasks. This may not be long enough to examine LTM

Explain: LTM often involves the consolidation of info over a more extended period, & a 20 minute break may not allow for adequate time for this process to occur. This may lead to an underestimation of the true capacity & duration of LTM

Link: Therefore, the results obtained in such experiments may not fully represent the extent to which LTM can store & retrieve information over extended periods

Limitation 2 of Research on coding - Acoustic STM?

Point: STM may not be exclusively acoustic

Evidence: Brandimote et al (1992) found that participants used visual coding in STM if they were given a visual task & prevented them from doing any verbal rehearsal in the retention interval before performing a visual recall task

Explain: Normally, we ‘translate’ visual images into verbal codes in STM, but as verbal rehearsal was prevented, participants used visual codes. Other research has shown that STM sometimes uses a semantic code (Wickens et al, 1976)

Link: This suggests that STM is not exclusively acoustic

Who developed Multi Store Model of Memory and what did they suggest about it?

• Atkinson & Shiffrin (1968)

• Suggested memory has 3 unitary stores which are linked by processes.

What are the 3 stores?

1. Sensory Register

2. Short term memory

3. Long term memory

What are the processes in the multi store model of memory?

1. information

2. Attention

3. Maintenance rehearsal

4. Transfer

5. Retrieval

6. Forgetting

What is the first step in remembering someting?

Attention

• If a person’s attention is focused on one of the sensory stores, then the data is transferred to STM

Sensory register

The place where info is held at each of the senses & the corresponding areas of the brain

• Constantly receives info but most of this receives no attention & remains in sensory register for a v. brief duration

Maintenance rehearsal

• Repetition keeps information in STM but eventually such repetition will create a LTM

• Atkinson & Shiffrin proposed a direct relationship between rehearsal in STM & strength of LTM

  • The more info is rehearsed, the better it is remembered so it can be transferred to LTM

Retrieval

The process of getting info from LTM, involves the info passing back through STM

• It is then available for use

Strength 1 of the Multi Store Memory Model - Supporting Evidence

Point: Research from brain scanning techniques has supported the Multi Store Memory Model & the idea of unitary memory stores

Evidence: Squire et al (1992) used brain scanning techniques & found that STM can be associated w. activity in the prefrontal cortex & that LTM can be associated w. activity in the hippocampus

Explain: Provides biological evidence that the diff types of memory are processed by different parts of the brain & that memory stores are distinct as the multi store model suggests

Link: Therefore, provides strong support for the 3 unitary stores proposed by MSM

Limitation 1 of the Multi Store Memory Model - Too simple

Point: The multi-store model suggests that both STM & LTM are single unitary stores. However, research doesn’t support this

Evidence: Working Memory Model & its supporting research shows that working memory (STM) is divided into a no. of qualitatively diff stores. Research shows there a no. of qualitatively kinds of LTM & each behaves differently.

Explain: For STM, there isn’t just a difference in terms of capacity & duration, but in the kind of memory stored there. For LTM, maintenance rehearsal can explain long-term storage in semantic memory but doesn’t explain long-term episodic memories

Link: This suggests that the MSM may be overly simplistic

Strength 2 of the Multi Store Memory Model - Supporting Case Studies

Point: A strength is that it is supported by the case of HM & shows that STM & LTM are unitary stores.

Evidence: HM’s new memories for LTM was badly damaged. He had no memory for events that happened just hours or even minutes earlier. His LTM never improved w. practice either. But testing showed his STM wasn’t affected as much at all. His almost normal digit span showed he could recall information that was presented to him immediately

Evidence: What happened to HM is evidence that it is possible to suffer damage to 1 of these stores w. the other remaining relatively unaffected.

Link: Therefore, this supports the notion of separate memory stores.

Limitation 2 of of the Multi Store Memory Model - Refuting Case Studies

Point:

Evidence: KF (Shalice & Warrington 1970) was in a motorcycle accident. He developed amnesia. When digits read aloud to him, his STM was poor- he had a digit span of 1. When he read digits to himself, his STM was better.

Explain: Suggests there is more than 1 STM store, 1 for auditory, & 1 for visual info.

Link: This challenges MSM’s idea of a single unitary STM store.