The multi-store model of memory

Atkinson + Shiffrin (1968). Called 'modal model' b/c for long time, most usually used model of memory. Called 'multi-store' b/c consists of 3 memory stores, linked to each other by processes that enable transfer of info from one store to next.

Place where infor held at each of senses eg eyes, ears, nose, fingers, tongue etc, + corresponding areas of brain.

Capacity very large. Constantly receiving infor, but most of this receives no attention + remains in sensory register for brief duration (milliseconds).

If attnetion focused on one of sensory stores, data transferred to STM. Attention 1st step in remembering.

Info is held in STM so can be used for immediate tasks eg working on maths problem. 

STM - limited duration - in 'fragile' state + will disappear (decay) quickly if not rehearsed. Maintenance rehearsal - repeating what you want to remember over and over again - largely verbal. 

Info also disappears from STM if new info enters STM, displacing original info. Happens b/c STM has limited capacity.

Repetition keeps info in STM but eventually repetition will create an LTM. Atkinson + Shiffrin proposed direct relationship b/ween rehearsal, better remembered - again called maintenance rehearsal.

Potentially unlimited duration + capacity. May feel many things you feel once knew but have forgotten, but evidence suggests either actually never made memory permanent, or it is there but can't find it - 'forgetting'.

Process of getting info from LTM involves info passing back through STM Then available for use.

Supporting evidence - controlled lab studies on capacity, duration + coding support existance of separate ST + LT store - basis of MSM. Studies using brain scanning techniques also demonstrated difference b/ween STM + LTM. Eg, Beardsley (1997) found prefrontal cortex active during STM but not LTM tasks. Squire et al (1992) also used brain scanning + found hippocampus active when LTM engaged.

Case studies - HM (Scoville + Milner, 1957). Brain damage caused by operation to remove hippocampus from both sides of brain to reduce severe epilepsy. HM's personality + intellect remained intact but couldn't form new LTMs, although couldn't remember thins from before surgery.

MSM too simple - suggests both STM + LTM single 'unitary' stores, but research doesn't support this. WMM suggests STM actually divided into number diff stores. Same true for LTM. Research shows number of qualitatively diff kinds of LTM + each behaves differently. Eg maintenance rehearsal can explain LT storage in semantic memory but doesn't explain LT episodic memories.

LT memory involves more than maintenance rehearsal - Craik + Lockhart (1972) suggested enduring memories created by processing you do, rather than through maintenance rehearsal; things processed more deeply more memorable just b/c of way they're processed. Craik + Tulving (1975) gave partitcipants list of nouns + asked q involved shallow/deep processing - asked whether word printed in capital letters (shallow processing) or asked whether word fitted into sentence (deep processing). Participants remembered more words in task w/ deep processing rather than shallow processing. 'Deep' processing key process in creating LT memories.