The multi-store model of memory

An information processing model of memory

Separate memory stores for sensory memory, short term memory and long term memory

Separate processes explaining how information passes between the stores and is either stored or forgotten

Multi store model of memory

Structure of the multi-store model
According to the multi-store model of memory (Atkinson & Shiffrin, 1968) memory can be explained in terms of 3 stores (sensory store, short term store and long term store) and 2 processes (attention and rehearsal).

Information first enters the sensory store (also known as sensory memory) directly from the senses. It remains in the sensory store for a maximum duration of around 2 seconds before it decays and is replaced with new information. If information in the sensory store is attended to then it can be passed to the short term store.

Around 7 plus or minus 2 chunks of Information (Miller, 1956) can be stored in the short term store (also know as short term memory). It is encoded primarily in a phonological format (by its sound) and remains there for around 12-30 seconds without being rehearsed. Where the information is rehearsed it can remain there for as long as it is being rehearsed, but it will stop any new information entering the store. Transfer from the short term store to the long term store is achieved by a process called elaborative rehearsal. New information that enters the short term store displaces (pushes out) any information that is already there, meaning that information that is not rehearsed and passed to the long term store is forgotten.

When information enters the long term store (also know as long term memory) it remains there for up to a life time (Bahrick et al, 1975). The capacity of the store is potentially unlimited, and encoding is primarily in a semantic format (information is stored by its meaning).


Evaluating the multi-store model
The multi-store model has fairly high face validity. This means that at first impression it seems to offer a plausible explanation as to how information passes through memory and is either stored or forgotten. Face validity however is only a small strength of the model, and what is needed to make a full and fair judgement of the validity of the model is empirical research evidence.

Evidence that there are separate short term and long term stores is provided by research into the primacy and recency effect (Glanzer & Cunitz, 1966). Participants were presented with a list of words one at a time, and asked to recall them either immediately or after a delay of 30 seconds - participants in the delayed recall group were prevented from rehearsing the words as they had to count backwards for the 30 seconds delay. The findings were that participants who recalled the list immediately remembered the first and last few words best, participants that recalled the list after a delay only remembered the first few words, and both groups of participants had difficulty recalling words from the middle of the list. The researchers concluded that words from the beginning of the list were remembered because they had entered long term memory (they called this the primacy effect), and words at the end of the list were remembered well in the immediate recall condition because they were still in short term memory (they called this the recency effect). The primacy effect and recency effect can be taken together to provide evidence that human memory has separate short and long term stores, and therefore provides evidence for the validity of the model as it too claims there are separate short term and long term memory stores.

Data from patients who have suffered brain damage leading to anterograde amnesia is another source of evidence for the validity of the multi store model. The most famous amnesia cases here are HM (Milner, 1966) and Clive Wearing. Both patients had severe damage to the areas of their brains that contain the hippocampi; HM’s brain was damaged during surgery to reduce epileptic fits, and Clive Wearing had a severe form of herpes that caused brain damage. After suffering their brain damage, both HM and Clive Wearing lost the ability to form new long term memories. Both had normally functioning short term memories, but as STM only has a duration of up to 30 seconds anything that happened to them was completely forgotten. They could remember things from their pasts prior to their brain damage. This severely debilitating condition provides strong neuropsychological evidence that short term and long term memory are completely separate entities in the human brain, and again supports the validity of the multi store model of memory.

So far then evidence from laboratory research into the primacy and recency effect, and evidence from brain damaged patients with anterograde amnesia has provided strong support for the multi store model. There are however cases that the multi store model cannot explain:

KF suffered brain damage in a motorcycle accident. Unlike HM and Clive Wearing, KF’s long term memory was normal, but his short term memory was damaged to the extent that he could only remember 2 items or chunks instead of the usual 7 + or - 2. This causes a problem for the model as it states that information must be retained and rehearsed in short term memory before it can be passed to long term memory - if KF’s short term capacity was reduced then following the model one would expect his ability to pass information to long term memory to also be reduced, but it was not and so the model must be flawed.

The famous case of anterograde amnesia discussed earlier, HM, also caused problems for the validity of the model when his long term memory was examined further. HM was completely unable to store new events in his long term memory, however he was able to learn new skills. One example of this was the pursuit rotor task in which he simply had to use a pencil to follow a wavering line on a rotating disc. Each time he was asked to repeat the task he was unable to remember ever having seen the disc before, but on each occasion he became more accurate through practice. HM had therefore learned a new skill, which must be stored in long term memory, and so the idea that long term memory is a single unitary store as represented by the multi store model must be incorrect; there must be more than one type of long term memory (e.g. declarative which was deficient in HM and non-declarative which remained intact).

Strengths: High face validity. Evidence for separate short and long term stores from research into the primacy and recency effect and anterograde amnesia.
Weaknesses: Cannot explain how KF is able to make new long term memories even though his short term memory is impaired. Does not account for there being more than one type of long term memory and the fact that only declarative LTM is damaged in anterograde amnesia patients.


A Level exam tips
How to answer a 12 mark question (PSYA1 AQA A specification)
Outline and evaluate the multi-store model of memory.
6 AO1 marks can easily be obtained by describing the model in quite simple terms. A drawing is perfectly acceptable and should be accompanied by a description of how information flows through the model and what happens to it. Remember to describe the essential processes of attention and rehearsal. Simple description of the stores should focus on their capacity, duration and encoding. Do not spend more than 10 minutes on this half of the answer.

6AO2 marks come from evaluating the model. Good advice would be to choose 3 evaluation points (e.g. primacy and recency, anterograde amnesia as support for the model, and KF) and describe them well making sure you explain exactly how they either support the model or are a problem for it. Again do not spend more than 10 minutes.