The multi-store model
The Atkinson and Shiffrin model - 1968
- Atkinson & Shiffrin envisaged memory as a flow of information through an information-processing system
- Information enters the system from the environment and first registers on the sensory memory store
- The short-term memory store has a very small capacity and information can be lost easily if not rehursed & transferred into long-term memory
- Once information reaches LTM it can stay for a lifetime, it can be lost through damage to the brain
- MSM structural model because it focuses on the storage components of the memory systemmost
- Important processes for MSM is rehearsal so information can be circulated
Sensory memory (SM) - Where stimuli first enter the memory
- SM holds information for fractions of a second after physical stimulus no longer available
Atkinson and Shiffrin proposed three separate sensory stores to accommodate different kinds of input:
- iconic store for visual input (things we see)
- echoic store for auditory input (things we hear)
- haptic store for tactile input (things we feel/touch).
- According to Baddeley (1988), purpose of the visual sensory store is to allow us to integrate visual information, at a conscious level, so we experience a smooth, continuous visual experience instead of a jumbled set of jerky, disconnected images. e.g. cartoons
Another possible function of SM is to sift through huge amounts of incoming sensory information. SM holds image of stimuli for a few milliseconds while scanned to decide which ones should be given attention & passed on through the system for further processing.
What is the evidence for sensory memory?
- Sperling (1960)
- Sperling used a chart containing three rows of letters, which he displayed for very brief exposures (50 milliseconds) to his participants
- Participants were immediately asked to recall as many of the letters as possible and could usually only recall about four or five, but frequently reported having been aware of more letters even though they could no longer recall them
- Trained participants to distinguish between three tones, participants were able to recall, on average, three items from whichever row had been cued by the tone
- Reason participants can only recall four items in the whole report technique is because the image of the words fades during the time it takes to report back these four items
- Stimuli used were artificial and may not reflect how we use memory in everyday circumstances
- No serious ethical issues, but fully informed consent must be gained
Differences between STM & LTM
- STM: Capacity & Duration = VERY LIMITED, Encoding = Maily ACOUSTIC, Forgetting = maily DISPLACEMENT
- LTM: Capacity & Duration = UNLIMITED, Encoding = Mainly SEMANTIC, Forgetting = mainly INTERFERANCE
Evidence for a distinction between STM and LTM
- Experiment: ask participants to study long lists of words and then to recall as many words as possible in any order (free recall).
- Found free recall of a list of unrelated words produces a characteristic serial position curve
- Words in the middle of the list have the lowest recall rate because they have had little time for rehearsal and have been displaced by later items in the list
- People can remember the last few words in the list (the recency effect) because the words are still circulating in the STM
- People can remember the last few words in the list (the recency effect) because the words are still circulating in the STM
- Later researchers argued that, if the different sections of the serial curve represent a genuine difference between STM and LTM, it should be possible to find ways of influencing one part of the curve but not the other
Research study: Glanzer and Cunitz (1966)
- Illustrates this so-called functional dissociation
- Their participants lists of words presented one at a time and then tested their free recall.
- 2 conditions in their experiment
- Condition 1, participants were asked to recall the words immediately after they had been presented - found the expected serial position curve
- Condition 2, participants were given a distractor task after the words had been presented and had to count backwards in threes for 30 seconds before they were asked to recall the words - found the expected serial position curve
- By suggesting that the task of counting backwards in threes had displaced the last few words in the list, but that the task had not affected the earlier words
- Highly controlled laboratory experiment and it has been replicated many times
- Artificiality of the task means that it might not represent how memory works in everyday life.
- Participants had to undertake several trials - done to avoid unrepresentative results
- Had to gain the informed consent of their participants and debrief them afterwards
- Strongest evidence for a distinction between STM and LTM come from the study of people who have suffered brain damage
- Several case studies of people with severe memory loss, including that of Clive Wearing
- HM suffered severe epilepsy, underwent brain surgery to remove parts of the temporal lobes & hippocampus
- Alleviated epilepsy but left with severe memory defecits (IQ remained same)
- STM appeared to be intact (could remeber 6 numbers in order)
- LTM damages, repeatedly read same magazinw withoug realising
- KF sustained brain injuries after a motorcycle accident had an intact LTM, he was able to learn new information & recall stored information.
- STM was affected so that he had a recency effect of only one item.
- STM & LTM have been shown as seperate stores through brain scanning
- Generally accepted the LTM has unlimited capacity
- Possible for info to be lost through processes such as decay & interferance but loss NOT because of capacity limitations
- LIMITED capacity
- Displacement of info common, e.g. trying to remeber a phone number, newer didgets tend to push out the first ones
- Rehursal (where you repeat things) helps to keep information repeating, if info trying to enter STM it breaks down rehursal process
- earliest systematic attempts to measure the capacity of STM was carried out in 1887 by Jacobs - the digit span technique
- Found people could remeber 7 digets or items (+/- 2)
- Miller thought that the chunk was the basic unit in STM and that we can recall 7 +/-2 chunks of information
- Other's questioned this assumption, claiming use of term 'chunk' is too vague
- Simon (1974), found that the span, measured in chunks, depends on the amount of information contained in the chunk
- Glanzer and Razel (1974) used the recency effect as a measure of the capacity of STM
- found that the recency effect was 2.2 items when the stimulus material consisted of single, unrelated words, increased to 1.5 sentences when unfamiliar sentences were presented and to 2.2 proverbs when familiar proverbs were used.
Factors that affect the capacity of STM
- Influence of LTM: Cowan believed that millar overstimated number of chunks that can be held in STM, thinks performance span task often affected by rehursal & LTM. Bower & Winzenz found diget srings replaced within series of immediate memory span trials became easier for participants to recall, suggests that have been gradually rehursed & stored in LTM, temporarily increasing capacity of STM.
- Reading aloud: digit span increases if participants read the digits aloud instead of reading them subvocally, suggests that this is because the digits are also then stored briefly in the echoic store which strengthens memory trace
- Pronunciation time: STM is determined by time constraints rather than structural limitations, found that the immediate memory span of young children was related to the length of time it took them to articulate words
- Individual differences: is some evidence that individual differences affect STM capacity
- STM: a temporary store and anything we need to retain for longer periods has to be transferred to LTM
- Repetition keeps the material in STM by continually reinserting it into the STM loop
- Rehearsal strengthens the memory trace so that it can be lodged permanently in LTM
- Research study: Peterson and Peterson
- presented participants with a consonant trigram, Participants were then asked to count backwards in threes from a specified number then repeat the trigram , this was repeated several times with different trigrams.
- Participants able to reall 80% of trigrams after 3sec interval (withoug rehursal) recall got gradually worse as time intervals lengthened.
- Issues: Possible loss of info was more to do with capacity limitations rather then duration.
Factors affecting duration (STM)
- Rehearsal: extend duration of STM by rehearsing information. e.g. look up a phone number in a directory, tend to repeat it over in our heads to hold it in our memory for long enough to dial it correctly.
- Intention to recall: seems to make a difference whether we are making a conscious effort to recall material. Sebrechts et al. (1989) tested serial recall for sets of three familiar English nouns. Condition where participants were not expecting to be asked to have to remember the words, correct recall fell to 1 per cent after--only four seconds.
- Amount information to be recalled: Murdock adopted Peterson & Peterson technique but used either a single, three-letter word or a set of three unrelated words. When used three words as stimulus, found same pattern of decline in recall as in the original Peterson & Peterson study. When used three letters (that formed a recognisable single word), recall was resistant to decay. Though rehearsal had been prevented, accurate recall level was about 90 per cent after 18 seconds. seems, that important factor is number of chunks rather than number of individual items.
Factors affecting duration (LTM)
- Experimental techniques: people seem able to remember things from distance past much better if given cues, rather than recall from scratch. Accuracy increases when measured by recognition rather than recall
- Depth of learning: people likely to remember things for longer if it has been learned well in 1st place.
- Pattern of learning: Bahrick found that people who had learned Spanish found vocab items learned over spaced sessions were retained for longer than vocab in an intensive session
- Nature of material to be learned: some types of material are retained for longer than others. Studies found that certain subjects recalled more accuratly over time.
- Info arrives in sensory mem it is still in origional form (visual image or sound)
- Sensory store has seperate stores for diff sensory experiences (modalities)
- Atkinson & Shiffrin = STM as unitary store (single storage space with no seperate compartments)
- Three main types of encoding used in memory:
- Aucoustic encoding: sound of stimulus
- Visual coding: physical apperacne of stimulus
- Semantic coding: meaning of stimulus]
- Research shown main type of encoding in STM = SOUND
- People are likely to confuse items that look sound similar if they are using acoustic code, where as they will confuse items that look similar if they are using visual code.
Research Study: CONRAD
- Participants shown a random sequence of 6 consonants, projected them rapidly in sequence onto a screen
- Two conditions:
- Letters ACOUSTICALLY SIMILAR
- Letters ACOUSTICALLY DISSIMILAR
- Immediatly after presentation participants asked to write letters down in correct serial order
- Conrad found that participants frequently made errors of recall, majority involving substitution of similar sounding letters
- Participants = difficult to recall strings of letters that sounded different
- Conrad concluded that we must convert visually presented material into acoustic code in STM & there is acoustic confusion.
- Methodological Issues:
- Ethical Issues: No serious ethical issues, but concent must be gained & participants debriefed.
Encoding (LTM) - Research Study: BADDELEY
- Constructed a pool of short, familiar words for each of four categories
- acoustically similar words
- acoustically dissimilar words
- semantically similar words
- semantically dissimilar words
- Each category presented random sequence of five words, asked participants to write down immediately after presentation (serial order)
- Found words that sounded similar were much harder to remember than words in any of the other categories
- Concluded, like Conrad, that STM codes acoustically.
- modified his experiment to test LTM.
- Extended length word lists from 5-10 & prevented participants rehearsing
- Each list presented 4 times, then recall tested after 20-minute interval found acoustic similarity = no effect on recall but words that were similar in meaning poorly recalled. LTM = semantically codes
Strengths of MSM
- Made many important contributions to memory research
- Information-processing approach enabled psychologists to construct testable modles of memory
- Many researchers would agree that there is a basic distinction between STM & LTM, there is pleanty evidence to support this distinction
Weeknesses of MSM
- Over simplified & fails to reflect complexity of human memory
- Considerable evidence that simple repetition = one of the least effective ways of transtering information
- Not a 1 way system, clear information from LTM must sometimes be activated before certain stages of processing can occur
- Much supporting evidence for MSM comes from lab experiments
- There is research results that MSM simply cannot explain.