Multi-store model of memory - created from Mind Map

The multi-store model of memory (Atkinson & Shiffrin) Representation of memory sees memory as transfer of info in 1 direction through 3 different stores.These 3 stores are: Each store differs in Capacity, Duration and Encoding Sensory information enters sensory memory. This info is held for fractions of seconds and is likely to decay if it's not attended to quickly. Atkinson & Shiffrin proposed separate stores within sensory store for each of the senses: Iconic-things we see. Echoic-things we hear. Haptic-things we touch. If info is sensory store is attended to it will be transferred to Short term memory store. Info in this store can remain if rehearsed but not for long. Info can be transferred to long term if it's rehearsed enough. It can be retrieved when necessary and can potentially remain there forever. Capacity of STM: Jacobs devised - 'The serial digit span'- findings: average STM span was between 5-9. Miller conducted review of studies that investigated capacity of STM which supported Jacobs study, conclusion: Capacity of STM- 7+/-2 chunks of info. Capacity of LTM In comparison to STM, research says capactiy of of LTM is unlimited. It is possible to lose data from LTM, Through processes like decay & interference, but forgetting does ot occur because of capacity limitations Duration of STM STM-Temporary memory store, anything that needs to be stored for a long time needs to be transferred to LTM. Brown-Peterson technique:ppts briefly shown consonant trigrams e.g. BVM & were given interference task of counting backwards in 3's to prevent rehearsal. After intervals of 3,6,9,12,15,18 seconds asked to recall trigram. Findings: No interference after 3 seconds but as time increased, performance decreased. After 18 seconds memory trace for trigram had decayed. Peterson & Peterson concluded duration a STM is about 18 seconds. Duration of LTM Some Psychologists argue LTM is for life. Certain forms of info have a longer duration. Like info based on skills-PROCEDURAL rather than facts-DECLARATIVE. Bahrick et al showed 400 ppts aged 17-74 yrs set of photos and list of names, some of which were old school friends , & were asked to identfy old school friends in the photos. Those who had lef school 48 yrs previously recalled 80% of names and 70% of faces suggesting memory of ames and faces is long lasting. Encoding Info arrives in it's original form, such as sound or vision. This is then encoded into a form that your memory can deal with This could be based on how it sounds- ACOUSTIC, what it looks lik VISUAL or what the work means SEMANTIC Encoding in LTM Encoding in STM Baddeley examined whether encoding in STM was primarily acoustic or semantic. 75 ppts presented with 1 of 4 word lists. List A acoustically similar. List B Acoustically disimilar. List C Semantically similar. List D semantically disimilar. Ppts given list containing original words and task was to arrange in correct order. Ppts given list A performed significantly worse with recall of 10% compared to other groups where recall was 60-80%. Appeared to be acoustic confusion with words in list A suggesting STM is encoded ACOUSTICALLY. Baddeley used same procedure as in STM study but with 20 minute interval before recall, during which ppts conducted another task. This time he found ppts that were given List C (semantically similar) only had 50% recall whereas other groups had 70-85%. As list C was recalled least effectively it appears there is semantic confusion in LTM suggesting it's encoded semantically.

Strengths of Multi store model Evidence for different stores- There is a wealth of evidence that demonstartes STM & LTM are seperate stores. CASE STUDY- Milner-attempted to treat patient's H.M. epilepsy by removing hippocampus. This resulted in H.M being unable to form new long term memories although his STM appeared to be unaffected. Because there was a difference in the way the 2 memory stores were affected. This shows that STM and LTM must be different stores. BRAIN SCANS- such as FMRI scans have shown that different areas of the brain are active when doing memory tasks. For example the prefrontal cortex is active when ppts are doing a STM task whereas hippocampus is active during LTM task. EXPERIMENTAL EVIDENCE- Glazer and Cunitz studied the serial position effect- if you give ppts list of 20 words then ask them to recall any words they can remember they tend to remember words from start of list and words from end (recency effect) The primacy effect happens because words at beginning have been rehearsed the most so transferred to LTM and recency effect occurs because these words are still in your STM when you begin recalling the list.

Weaknesses of Multi store model STM & LTM ARE NOT UNITARY STORES- Model is regarded as an oversimplification of structures involved in memory. It suggests that btoh LTM and Stm are unitary stores, but research suggests this is not the case: REHEARSAL NOT ONLY PROCESS- There is a considerable evidence that simple repetition is one of the least effective ways of remembering. INFO DOESN'T FLOW IN ONE DIRECTION STUDIES NOT VALID- Much of research for this model comes from artificial lab experiments which are well controlled but may not reflect how memory works in everyday life. Furthermore, memory experiments tend to use words and letters which are only relevant to some aspects of memory. The experiments also tend to use psychology students so conclusions may not apply to everyone. Clive Wearing lost his LTM episodic memory but not procedural memory suggesing there may be more than one type of LTM Craik & Lockhart suggested enduring memories are created by the processing you do rather than through maintenance rehearsal. The more deeply you process something, the more likely you are to remember it. In an experiment ppts shown list of nouns & asked question about each word. This was either 1) Shallow processing 2)Phonemic processing 3) Deeper processing. Findings: Ppts remembered more from condition 3 and the least from 1 suggesting deeper processing of material leads to enhanced memory. This has led to proposal of elaborative rehearsal as away of transferring info from STM to LTM Model implies only new info enters STM. However this incorrectly represents the processes of memory- it is not possible to make sense of new info without drawing on info already stored in LTM. Evidence for this comes from study by Groot who found that STM for the position of chess pieces on a board was much better in chess-players than non-players.

Sorry, but you may only upload 10 images per page.

Multi-store model of memory