The Multi-store Model of Memory

Duration: how long a memory lasts for before it is no longer availiable.

Encoding: The way information is changed so that it can be stored in memory

Capacity: A measure of how much can be held in memory. Measured in terms of bits of information.

SM

• Duration: milliseconds
• Encoding: depends on sense organ involved
• Capacity: all sensory experience

STM

• Duration: 20 seconds
• Encoding: mainly acoustic
• Capacity: 7+/-2 bits

LTM

• Duration: potentially unlimited (anywhere from 2 minutes - 100 years)
• Encoding: mainly semantic
• Capacity: very large

• Information picked up by sensory organs in the SM store
• Disappears unless passed onto the STM store
• Pay 'attention' in order to enter STM
• 'Maintenance rehearsal' to transfer information into the LTM

Aim: to find evidence to indicate the duration of the sensory store.

• Participants showed a grid of digits and letters for a blink of an eye.
• Either asked to write down all 12 items or told they would hear a tone and immediately after hearing the tone, they should write down just that row.

Results: Their recall was poorer (5 items recalled - 42%) than when asked to give a row only (3 items recalled - 75%).

Conclusion: This shows that infomation decays rapidly in the sensory store

Aim: To investigate the duration of short-term memory, and provide empirical evidence for the multi-store model.

• 24 participants had to recall trigrams (meaningless three-consonant syllables, e.g. TGH).  To prevent rehearsal participants were asked to count backwards in threes or fours from a specified random number until they saw a red light appear. This is known as the brown peterson technique.
• Participants were asked to recall trigrams after intervals of 3, 6, 9, 12, 15 or 18 seconds.

Results: The longer the interval delay the less trigrams were recalled. Participants were able to recall 80% of trigrams after a 3 seconds delay. However, after 18 seconds less than 10% of trigrams were recalled correctly.

Conclusion: Short-term memory has a limited duration when rehearsal is prevented. It is thought that this information is lost from short-term memory from trace decay. The results of the study also show the short-term memory is different from long-term memory in terms of duration. Thus supporting the multi-store model of memory.

Shepard (1967)

He showed participants 612 memorable pictures one at a time. An hour later they were shown these pictures again and showed almost perfect recognition. Four months later, they were still able to recognise 50% of the photographs.

Bahrick et al (1975)

 He asked people of various ages to put names to the faces in their high school year book. 48 years on and people were 70% accurate.

There are a string of numbers and you read the 1st numbers, then shut your eyes and recall it. Next you try it with two digits and again shut your eyes and recall it. You repeat this and see how many numbers you can remember.

Miller (1956) concluded that the average span for digits was 9.3 while it was 7.3 for letters. He suggests this may be because there are only 9 digits whereas there are 26 letters.

The size of the chunk affects how much you can remember because it was found that people who had a shorter memory span for larger chunks than smaller chunks.

Individual Differences

Jacobs found that the digit span increased steadily with age. In one sample of a school of 8 year old girls, they could remember an average of 6.6 digits whereas the average for 19 year olds was 8.6 digits.

The digit span might increase with age due to a gradual increase in brain capacity or people develope strategies to improve their digit span as they get older such as chunking. (A process of taking individual units of info and grouping them into larger units e.g. 4-7-1-1-3-2-4 : 471-1324)

• viral infection, attacks brain, damages the hippocampus
• lost ability to transfer memories from STM to LTM
• see's what is in front of him but it fades once he looks away
• repeats himself
• always thinks he's just woken up every two minutes
• same entry throughout his diary
• thinks he was unconscious when he wrote previous entries so he crosses them out
• whenever he see's his wife, it's like he's seen her for the first time in years
• reading, writing and playing music is automatic for him

Glanzer and Cunitz (1966)

• This study showed that if you give participants a list of 20 words, one at a time, and then ask them to recall the words they remember, there is an interesting effect.
• They tend to remember words from the start of the list (primacy effect) and from the end of the list (recency effect) but words in between are not as well remembered.
• The primacy effect occurs because the first words are being rehearsed and transferred into the LTM. The recency effect occurs because the words are in the STM when you start recalling the list. Other words that aren't at the beginning or at the end of the list may be reaclled because you remember that word by association.

How do primacy and recency effects support the distinction between STM and LTM?

The primacy effect shows us that the first word has gone into our STM store and has been rehearsed more than other words so goes into our LTM helping us to be able to remember it better when recalling. The recency effect shows us that the last words have entered our STM and we are able to remember the last words because they are in our STM and it's the last thing we recalled. 

The primacy supports the STM and the recency supports the LTM

Strengths

• provides evidence to support the distinction between STM and LTM
• the model can account for primacy and recency effects
• the model is influential as it has generated a lot of research into memory
• basis of MSM is sound 
• three qualitively different stoes: echoic, haptic and iconic
• model provides account of memory in terms of structure and process
• MSM had clear predictions about memory so psychologists can conduct studies to test it

Weaknesses

• it oversimplfies memory structures and processes
• structures: STM and LTM are not unitary stores
• MSM suggests STM and LTM each operate but they actually work together
• MSM suggests LTM as one single store - Schachter et al suggests there are four LT stores

The Four Long-term memory stores Schachter et al suggests there are

• Semantic Memory: memory for knowledge about the world
• Episodic Memory: memory for what you did yesterday or a film you saw last week
• Procedural Memory: memory for riding a bike or learning how to read
• Perceptual Representation System (PRS): enhanced recognition to specific stimuli which have been seen before (e.g. words)

Processes: Rehearsal VS Processing

• Maintencance rehearsal is not the only way LTM are created
• Craik and Lockhart (1972) suggested that enduring memories are created by the processing that you do, rather than maintenance rehearsal

How separate are STM and LTM?

• MSM suggests STM is involved before LTM
• Logie (1999) pointed out that STM relies on LTM so cannot come first as suggested
• Ruchkin et al (2003) concluded that STM is just the part of LTM which is activated at any given time
• MSM's STM was a place to hold info as it is being moved into the LTM - whereas concept of working memory is it's a place where info is held until it is put use.

Atkinson and Shiffrin (1968) proposed that memory could be explained in terms of three differentstores: sensory memory (SM), short-term memory (STM) and long-term memory (LTM).

The SM store has three other parts within the store- the echoic store, the haptic store and the iconic store. Our sensory memory picks up information from these three stores however once the information is picked up by our sensory organs it decays in seconds. SM has a duration (how long a memory lasts before it is no longer availiable) of milliseconds but if a person's attention is focused, it will then be transferred into the STM. It has a limited capacity (a measure of how much can be held in memory).

STM has a a duration of 20 seconds and it's capacity is 7+/-2 bits. STM's encoding (the way information is changed so that it can be stored in memory) is mainly acoustic however it also has visual encoding. 

The LTM store has a potentially unlimited duration, has semantic encoding and has a very large capacity. Attkinson and Shiffrin claimed that memories from STM can be transferred into the LTM by rehearsal. The more information is rehearsed, the better it is remembered.