The multi-store model of memory: include the capacity, duration and coding of each store. - Memory Psychology

THE MULTI-STORE MODEL OF MEMORY

It was proposed by Atkinson & Shiffrin (1968) and it attempts to explain how memory works, in terms of 3 stores: the sensory register, short-term memory, and long-term memory.

Coding is…the means by which information is represented in memory. For example: if you can picture the information, you’ve used visual coding; acoustic coding uses sound and sematic coding used meaning.

Capacity is…the amount of information held in a memory store.

Duration is…the length of time information can be held in memory.

Short-term memory has a limited capacity of between 5-9 items. It is the information that we are currently aware of and can process and recall straight away. The coding used is usually verbal (acoustic), and we have to pay attention to new information for it to be in our short-term memory. STM has a duration of 18-30 seconds.

Long-term memory has an unlimited capacity, and a potential duration of a lifetime. This is a permanent memory store and the information is not conscious, but can be recalled when needed. The coding used is usually semantic. If we rehearse information from our STM, it can be transferred to our LTM.

---------------

Research into coding:
BADDELEY (1966)
Aim: to explore the effects of acoustic and semantic coding in short-term and long-term memory.

Procedure: it was a lab experiment with four conditions: acoustically (dis)similar words and semantically (dis)similar words. The participants were asked to recall the words either immediately (STM) or after a delayed period of time (LTM).

With immediate recall, participants did worse with acoustically similar words suggesting that the coding in STM is acoustic. And, long-term memory appears to be semantically encoded, due to the more errors being made in the semantically similar condition when there was a delay.

These findings suggest that the nature of encoding is different for LTM and STM.

Evaluation: the findings of this study have limited application to real life as Baddeley used artificial stimuli, opposed to material which carried personal meaning to the participants. Because of this, we need to be cautious about generalising these findings to different memory tasks (ie. meaningful material may make us use a different type of coding).

---------------

Research into capacity:
JACOBS (1887): ‘digit span’ test

Aim: to investigate the capacity of short term memory.

Procedure: a sample of 443 female students (aged 8-19) from a North London school took part. They had to repeat back a string of numbers or letters in the same order (the amount of digits/letters were gradually increased until they were no longer able to recall the sequence).

Findings: the mean recall for digits was 9.3 and the mean recall for letters was 7.3

Conclusion: this supports Miller’s notion of 7±2.

Evaluation: this study is very old. Early psychological research didn’t have the same levels of control that modern psychology now has, and therefore, there may have been confounding variables (such as, the participants being distracted, which could’ve affected the results), meaning this study lacks internal validity. In contrast, the results of this study have also been found by other research (eg. Miller (1956)), which supports the validity of this research.

-----------------------

Millers Magic Number (1956):
Miller observed everyday life and noted that things often happen in 7’s (eg. 7 days in a week, 7 deadly sins) – which led to his suggestion that the capacity of STM is  7±2.

He also noted that people can recall 5 words just as well as they can recall 5 letters – done through a process of ‘chunking’.

Chunking is…grouping sets of digits or letters into ‘units’ or ‘chunks’.

One limitation of Miller’s research is that he could’ve overestimated the capacity of the STM – Cowan (2001) reviewed other research and found that the capacity of the STM was only  4 chunks – suggesting that Miller’s lower estimate was more accurate.

------------------

Research into duration:
PETERSON AND PETERSON (1959)
Aim: to investigate the duration of short-term memory

Procedure: 24 university students were presented with a 3-letter nonsense syllable (‘trigram’) – and were asked to count backwards in 3’s from a specific number, to prevent rehearsal of the letters. Each participant was given 8 trials and were asked to recall the trigram after 3,6,9,12,15 or 18 seconds of counting backwards.

Findings: the longer they spent counting backwards, the less trigrams were recalled. Participants recalled 90% of trigrams correctly after a 3-second interval, but only 2% correctly after an 18 second interval.

Conclusion: This suggests that STM has a very short duration unless rehearsed. The results also show that STM is different from LTM in terms of duration and this therefore supports the multi-store model of memory.

Evaluation: a limitation is that the stimuli was totally meaningless and trying to memorise nonsense trigrams doesn’t reflect real life activities and thus, this research lacks external validity.

Bahrick et al (1975):
Aim: to test the duration of long term memory
Procedure: he studied 392 participants from Ohio (aged 17-74) and tested their recall using photo recognition (using 50 pictures from their high-school yearbooks), and a free-recall test (participants attempted to remember the names of people in their graduating class).

Findings: participants tested within 15 years of graduation were 90% accurate in photo recognition and after 48 years, they were 70% accurate. Free recall was lower.

Conclusion: long-term memory can last for a very long time.

Evaluation: his study has high external validity as real-life, meaningful memories were studied, and therefore these results reflect our memory for real-life events, and so, can be applied to everyday human memory.

However, this study also carries limitations as confounding variables are not controlled for: for example, participants may have looked at their yearbook photos over the years and rehearsed their memories – hereby weakening the validity of this research.

----------------

The sensory register has a very brief duration of ½ of a second but quite a large capacity – and it has a different store for each of our senses (‘sensory stores’).

Duration of Sensory Register:
Crowder (1993) found that the sensory register only holds information in the iconic store for a few milliseconds, but 2-3 seconds within the echoic store. (this supports the idea of sensory information being coded into different sensory stores).

-------------

EVALUATION:

P: large base of supporting research
E: for example, Baddeley (1966) found that the nature of encoding is different for STM and LTM; Peterson and Peterson (1959) found that the duration was different for STM and LTM, and HM had his LTM impaired during surgery but his STM remained intact – suggesting that short and long term memory are in separate areas of the brain.

E: with so much supporting evidence, it suggests that the MSM has very high validity

P: we form memories to useful things (eg. people’s faces, places, ect.), but a large majority of research used to support the MSM didn’t use such materials

E: most research used artificial stimuli (eg. Peterson and Peterson (1959) used non-sense trigrams).

E: this means that the MSM might not be a good explanation for how memory works in our everyday lives and therefore it lacks validity.

P: it has been criticised for being too simplistic

E: it assumes that STM and LTM are individual stores made up of one component each – but other research suggests that they consist of multiple components (eg. LTM has episodic, procedural and sematic memory).

E: therefore, the validity needs to be questioned

P: Shallice and Warrington (1970) had a brain-damaged patient, KF, who impaired his STM during a motorcycling accident.
E: However, he was still able to process and store infomation into his Long-Term memory
E: Whilst this does support the idea of there being separate stores, the MSM states that information must be passed from the STM into the LTM to store it – how could this be possible if his STM was impaired? Therefore, this theory has questionable validity.