Multi-Store Model of Memory (Atkinson and Shiffrin)
The multi-store model of memory suggests that memory is a linear process where memory has three different stores. These are distinguished in terms of capacity, duration and how memories are encoded. The first stage of memory is sensory memory (SM). This is a very brief stage where information comes from the senses and lasts up to 2 seconds. If attended to, then it is passed onto short term memory (STM) or if not attended to, it is lost. STM lasts around 15 – 30 seconds and needs rehearsal in order to maintain it. Information in STM is mostly phonetic. STM has a capacity of 5 – 9 items (Miller) and this can be increased through chunking. If information is rehearsed in STM, it is passed onto long term memory (LTM). Information not rehearsed is lost from STM, this can occur through displacement (not being able to store more information) or decay. LTM is potentially an unlimited store of information with an unlimited duration (minutes to years). Information stored in LTM is mostly semantic, and some phonetic and visual. Information may be lost from LTM through decay or interference. It is retrieved through use of retrieval cues.
Supporting Evidence:
The serial position curve shows that we remember words at the beginning and end of a list better than words in the middle. This is due to words at the beginning being rehearsed and converting to LTM and words at the end of the list being the most recent and in our STM. This suggests memory has different stores. Glanzer and Cunitz study supports this.
However this research is lab based and requires pp’s to recall a list of words. This is an artificial task and lack ecological validity, as it is not an everyday task to be asked to recall word lists.
Opposing evidence:
This model does not take into account different things we have to remember e.g. doesn’t consider interesting, relevant, funny – it treats all memories the same. It cannot explain how flashbulb memories occur
Craik & Lockhart suggest repetition is one of the worse ways of retaining information, yet the multi-Store model emphasises role of rehearsal.
The levels of processing can also explain the primacy and recency effect, meaning this evidence is not specific to the multi-store model. Craik and Lockhart say that the recency effect occurs because words are in consciousness and use maintenance rehearsal, while the primacy effect is due to rehearsal and deep processing as pp’s know they have to learn the words.
Different explanation:
The Levels of Processing model of memory suggests it is the depth material is learnt that leads to LTM – information which requires deeper processing will lead to better memory than information that requires shallow processing
Application:
Brain damaged patients provide strong evidence for the STM and LTM distinction, as brain damage can affect one store and not the other e.g. Clive Wearing.
However Multi-Store Model suggests memory is linear - how can some have impaired STM but LTM still fully functioning? Other cases have found patients with brain damage to STM to be able to add to LTM. KF had a motorcycle accident that affected his STM, however it only affected it verbally and he could still process visual information (Shallice and Warrington)
The multi-store model of memory suggests that memory is a linear process where memory has three different stores. These are distinguished in terms of capacity, duration and how memories are encoded. The first stage of memory is sensory memory (SM). This is a very brief stage where information comes from the senses and lasts up to 2 seconds. If attended to, then it is passed onto short term memory (STM) or if not attended to, it is lost. STM lasts around 15 – 30 seconds and needs rehearsal in order to maintain it. Information in STM is mostly phonetic. STM has a capacity of 5 – 9 items (Miller) and this can be increased through chunking. If information is rehearsed in STM, it is passed onto long term memory (LTM). Information not rehearsed is lost from STM, this can occur through displacement (not being able to store more information) or decay. LTM is potentially an unlimited store of information with an unlimited duration (minutes to years). Information stored in LTM is mostly semantic, and some phonetic and visual. Information may be lost from LTM through decay or interference. It is retrieved through use of retrieval cues.
Supporting Evidence:
The serial position curve shows that we remember words at the beginning and end of a list better than words in the middle. This is due to words at the beginning being rehearsed and converting to LTM and words at the end of the list being the most recent and in our STM. This suggests memory has different stores. Glanzer and Cunitz study supports this.
However this research is lab based and requires pp’s to recall a list of words. This is an artificial task and lack ecological validity, as it is not an everyday task to be asked to recall word lists.
Opposing evidence:
This model does not take into account different things we have to remember e.g. doesn’t consider interesting, relevant, funny – it treats all memories the same. It cannot explain how flashbulb memories occur
Craik & Lockhart suggest repetition is one of the worse ways of retaining information, yet the multi-Store model emphasises role of rehearsal.
The levels of processing can also explain the primacy and recency effect, meaning this evidence is not specific to the multi-store model. Craik and Lockhart say that the recency effect occurs because words are in consciousness and use maintenance rehearsal, while the primacy effect is due to rehearsal and deep processing as pp’s know they have to learn the words.
Different explanation:
The Levels of Processing model of memory suggests it is the depth material is learnt that leads to LTM – information which requires deeper processing will lead to better memory than information that requires shallow processing
Application:
Brain damaged patients provide strong evidence for the STM and LTM distinction, as brain damage can affect one store and not the other e.g. Clive Wearing.
However Multi-Store Model suggests memory is linear - how can some have impaired STM but LTM still fully functioning? Other cases have found patients with brain damage to STM to be able to add to LTM. KF had a motorcycle accident that affected his STM, however it only affected it verbally and he could still process visual information (Shallice and Warrington)