multi store model

The multi-store model provides a general overview of the workings of our whole memory system. It was created by Atkinson and Shiffrin.

Memory stores have different:

• Encoding (form of which info is stored in)

• Capacity (how much info can be stored)

• Duration (how long info can be stored for)

sensory register

Information gathered by sense organs enters sensory memory, only a small amount paid attention to passes to short term memory and the rest is quickly lost.

• Encoding —> raw information sorted into the 5 sensory stores: iconic, echoic, haptic, gustatory and olfactory

  • Crowder

    • Found that SR only retains info in the iconic store for a few milliseconds but for 2-3 seconds within echoic

    • Supports the idea that sensory info is coded in different ways

• Capacity —> large information contained in an unprocessed, highly detailed and ever-changing format

  • Sperling

    • Overall recall for each row was high (75% correct) which indicates that info was originally stored in SM

    • Therefore, showed capacity of iconic store is about 12 items and fades quickly

• Duration —> all stores have relatively limited duration through each store is not constant (decays at different rate), some evidence that duration decreases with age

  • Walsh and Thompson

    • Iconic sensory store has an average duration of ½ second, which decreases as individuals get older

    • This suggests duration of sensory memory is limited and dependant on age

short term memory

Information transferred to LTM for more permeant storage if in ST memory it is actively processed enough (thought about) through rehearsal. STM is a constantly changing (active) store, containing info currently thought about.

• Encoding —> info received from SR in its raw form, STM encodes it in a way that is easy to deal with:

  • Visually — thinking of the image

  • Acoustically — repeatedly saying the word (most common)

  • Semantically — using knowledge of word and its meaning (schema)

  • Baddeley

    • Found that people recall things via acoustics, assumed this because there was a confusion with acoustically similar words

• Capacity —> limited, research indicates between 5-9 items can be held but can be increased through chunking (the size of the units of information in storage is increased by giving them a collective meaning)

  • Jacobs

    • Serial digit span method used, presented w/ increasingly long lists of numbers to recall in right order

    • Ppt fail on 50% = said to have reached capacity

    • Found average memory span was between 5 and 9 items (digits - 9, letters - 7 as there are fewer numbers)

• Duration —> maximum of 30 seconds, can be extended longer if it is rehearsed. If done for long enough then info transfers to LTM (elaborative rehearsal)

  • Peterson and Peterson

    • Nonsense trigrams (don’t make words), asked to count backwards in 3s from 3 digit number to stop them rehearsing

    • Ppts able to recall 90% of trigrams after 3 seconds, only 5% could recall all correctly after 18 seconds —> concluded STM max duration between 18-30 seconds

long term memory

Information rehearsed more is transferred into LTM (elaborative rehearsal), the LTM stores info over long periods (over 30 seconds) and is not passive (changes over time, causing some memories to be inaccurate)

• Encoding —> occurs through meaning, although some research suggests coding can occur visually and acoustically

  • Baddeley

• Capacity —> potential capacity is unlimited, info can be lost due to decay and interference but losses don’t occur due to capacity

  • Anokhin

    • Estimated the number of possible neuronal connections in the neuron is 1 followed by 10.5km of 0s

    • He concluded that no human yet exists who can use all the potential of their brain, suggesting capacity is unlimited

• Duration —> depends on lifespan, as memories can last a lifetime they have a longer duration if originally well coded, and memories for skills have longer duration. Info in LTM does not have to be continually rehearsed to be obtained

  • Bahrick et al

    • 400 ppts 17-74 given photos and a list of names, some of which were ex-school friends, and asked to identify them

    • Left school in last 15 years: 90% recall accurate, left last 48 years: 80% names and 70% faces — suggesting LTM duration is a long time

baddeley study

• Aim: investigate whether acoustic similarity and semantic similarity influence short term and long term memory differently

• Procedure: ppts given lists of words that were acoustically similar (e.g cat, mat, sat) and semantically similar (e.g big, large). They were asked to recall immediately (STM) or after a delay (LTM)

• Results: ppts remember words by sound in STM and meaning in LTM (as they remembered these worst)

• Conclusion: different encoding preferences in STM and LTM

evaluations of the MSM

• MSM is supported by amnesia cases. In an attempt to treat patient HM’s epilepsy, removed brain tissue related in him developing anterograde amnesia, being unable to encode new long-term memories, although his STM was intact — supporting MSM idea that we have separate memory stores

• MSM supported by Murdock’s research. Found that ppts remember first and last few words but forget the middle ones in the list. Suggested that early words were in LTM (primacy effect) because they have time to rehearse each word acoustically and words at the end are still in STM (recency effect). This supports MSM idea that there are separate memory stores.

• This research used to investigate the duration, coding and capacities within each store of the multistore model lacks ecological validity. The studies were all laboratory experiments which mean they took place in an artificial setting with a highly controlled environment. Ppts were asked to take part in unrealistic tasks such as Peterson and Peterson’s research where ppts had to remember nonsense trigrams whilst counting backwards in 3s from a 3 digit number. Therefore, due to this artificiality, this task does not reflect real life, meaning we cannot generalise our results to other settings, specifically natural settings.

• The main criticism for the MSM is that it is oversimplified. The MSM assumes there is only one store for STM and LTM. As well as research investigating different types of LTM such as procedural and semantic, research conducted by Shallice and Wallace investigated the case study of KF who was brain damaged from a motorcycle incident. KF’s short term memory was severely impaired for auditory information, only being able to hold 2 digits when told but his STM for visual info seemed unaffected. This suggests that there is more than one type of short term memory, which is supported and explained by Baddeley and Hitch using their WMM to explain different types of short term memory, suggesting MSM is oversimplified