Architecture of memory
Several memory theorists (i.e. Atkinson & Shiffrin, 1968) proposed 3 types of memory store.
1. Sensory stores each hold very brief information. This is modality specific (limited to one of the five sensory modalities.)
2. STM of limited capacity ( - / + 7)
3. LTM of unlimited capacity holding information over extremely long periods of time.
Environmental information is initially received by the sensory stores. Information is held very briefly in the sensory stores, with some being attended to and processed further by the STM. Some information processed by the STM is transferred into the LTM.
LTM storage depends on rehearsal, with a direct relationship between the amount of rehearsal and the strength of the stored memory trace.
Our senses are constantly bombarded with information, most of which does not receive any attention. Info. in every sensory modality persists briefly after the end of the stimulation so that it can potentially be further analysed.
Iconic store (visual)
Classic work in this area was conducted by Sperling (1960). He presented a visual array containing 3 rows of four letters each for 50 ms to participants. Ps reported 4 - 5 letters but claimed to have seen many more letters. Sperling suggested that this happened because visual information had faded before most of it could be reported. He tested this by asking Ps to recall only part of the info. He found that visual store lasts around 0.5 seconds
Echoic store (audio)
This store holds relatively unprocessed audio input. Treisman (1964) asked people to shadow a message heard in one ear whilst ignoring the message in another. When the non-shadowed message preceded the shadowed one, the 2 messages were only recognised as being the same when they were within 2 seconds of each other. This suggests the temporal duration of unattended audio material is 2 seconds.
Short term memory
Trying to remember a telephone number in an everyday example of the use of our STM. The STM has 2 characteristics -
1. It has a very limited capacity (7 + / - 2)
2. It is fragile
The capacity of the STM has been assessed by span measures and by the recency effect in free recall. Digit span involves participants repeating back random digits in the correct order which they've heard them. The recency effect in free recall refers to the finding that the last few items in a list are usually better remembered in immediate recall than the items in the middle of a list (Glanzer & Cunitz, 1996)
The distinction between short and long term memory
Strong evidence for 2 distinct stores comes from studies with brain damaged patients. Two tasks probably involve different processing mechanisms if there is a double dissociation. Amnesic patients have generally poor long term memory but intact short term memory, for example. The reverse problem is relatively rare but does exist (i.e. the case study of KF, reported by Shallice & Warrington, 1970 - he had a recency effect of only 1 item but an intact LTM. However, he did not perform badly on all STM tasks. more later!)
Petersen & Petersen (1959) studied the duration of STM by using the task of remembering a 3 letter stimulus for a few seconds while counting backwards in 3s. The ability to remember the stimuli lasted only about 6 seconds, suggesting that information is lost rapidly from the STM.
Why count backwards? This acts as a source of interference, or it may divert attention away from info. in the STM. This is not the same as forgetting..
Evaluation of the multi-store model
The three memory stores differ in the following ways -
1. temporal duration
2. storage capacity
3. forgetting mechanisms
4. effects of brain damage
Many contemporary memory theorists have used the MSM as the starting point of their theories. Much theoretical effort has gone into providing a more detailed account of the LTM than is offered in the original by Atkinson & Shiffrin (1968).
However, the MSM as it is is oversimplified. It assumes that both the STM and LTM are unitary and always operate in a single, uniform way. Conflicting evidence was found from the case study of KF (Shallice & Warrington, 1972) - his memory deficit only concerned verbal materials such as letters, words etc. and did not extend to meaningful sounds (i.e. telephones ringing). So, KF's problems were not just STM as a whole but what the researchers termed his 'auditory-verbal short-term store.'
Another problem with the MSM is that LTM is too simplified. We hold so much information in that area that it's improbable that its stored in one single store.
Also, it assumes that information has to go through the STM before reaching the LTM, but that isn't what happens. For example, our ability to engage in verbal rehearsal of visually presented words depends on prior contact with stored info. concerning the words, i.e. pronunciation (Logie, 1999.)
One of the key components in the MSM was the short term memory (STM) store. This has since been revised by many theorists, resulting in the standard model (Nairne, 2002). These are the new, revised STMs assumptions..
1. Information in STM is in a state of activation
2. 'Permanent knowledge is activated' by the presence of this information (Nairne, 2002)
3. Currently activated information can be accessed immediately and effortlessly
4. Activation is fragile and can quickly decay (Nairne, again)
activated info. from the LTM is in the STM and decay causes that info. to leave the STM. Decay can be prevented by rehearsal.