THE MULTISTORE MODEL OF MEMOR
A SUMMARY OF THE DIFFERENCES BETWEEN SM STM AND LTM
CAPACITY
SENSORY MEMORY: VERY LARGE (SPERLING)
SHORT TERM MEMORY: 7+0R -2 ( MILLER, JACOBS, BADDELEY)
LONG TERM MEMORY: INFINITE (ANOKHIN)
DURATION
SENSORY MEMORY: APPROXIMATELY 200-500 milliseconds, DEPENDING ON SENSE MODALITY (CROWDER)
SHORT TERM MEMORY: 18-30 SECONDS (PETERSON &PETERSON)
LONG TERM MEMORY: INFINITE (BAHRICK)
ENCODING
SENSORY MEMORY: ICONIC, HAPTIC, ECHOIC, OLFACTORY, GUSTATORY (TRIESMAN)
SHORT TERM MEMORY: ACOUSTIC AND VISUAL (BADDELEY)
LONG-TERM MEMORY: MOSTLY SEMANTIC BUT ALSO VISUAL AND ACOUSTIC (BADDELEY)
THE MULTI-STORE MODEL OF MEMORY BY ATKINSON AND SHIFFRIN (1968)
.
A01/OUTLINE CONTENT
The specification says you must know: The multi-store model of memory: sensory register, short-term memory and long-term memory—features of each store: coding, capacity and duration.
Keywords:
Sensory Register/Memory (Iconic, Echoic, Haptic, Olfactory, Gustatory)
Short-Term Memory
Long Term Memory
Trace Decay
Transfer of STM to LTM
Forgetting Through Retrieval Failure
Linear Direction
Forgetting Through Displacement
Unitary Store
Multiple Stores
Rehearsal Loop
Encoding
Capacity
Duration
Acoustic Encoding
Semantic Encoding
Visual Encoding
MARK SCHEME FOR A01:
Atkinson and Shiffrin’s (1968) multi-store model of memory (MSM) distinguishes between the separate stores of sensory, short-term and long-term memory.
Likely features include:
It is a structural model
STM and LTM are unitary stores
Information passes from store to store in a linear way
Rehearsal is needed to pass information from STM to LTM
Each store has its characteristics in terms of encoding, capacity and duration
Explanations for forgetting are different for each store.
Candidates may include a diagram. If accurately labelled and sufficiently detailed, this can potentially receive the full six marks.
OUTLINE ESSAY EXAMPLE (A01)
Before the 1960s, little was known about cognitive processes such as memory. This is because the cognitive revolution was still in its infancy, and the behaviourist tradition that dominated the first half of the 20th century paid no regard to internal processes, believing that the brain, or ‘black box’ as they called it, was irrelevant.
It was assumed that memory was a unitary store in the brain.
Atkinson & Shiffrin (A&S) (1968) proposed that memory was not a singular store but contained three multi-stores, hence the model’s name, the multi-store model of memory. The three stores were sensory memory (SM), short-term memory (STM) and long-term memory (LTM). A&S also proposed that memory was a structural system rather than a process system and that information (or memory) linearly passed between the three stores; for example, SM must first be processed before information could pass to STM and then finally to LTM.
Sensory memory is the first storage system for incoming information and has several sensory registers (SR), e.g., iconic, echoic, haptic, gustatory and olfactory. Each of these registers processes incoming data from one of these five senses. Information received in SM is raw and unprocessed and has a large capacity. The duration of storage, however, is milliseconds, so unless given focused attention, it will fade from SM, never to be retrieved.
Sensory memory is a mental representation of how a person’s environmental events look, sound, feel, smell, and taste; e.g., people do not necessarily pay attention or notice the sounds and sights they continually encounter when conscious. Most sensory data goes unnoticed; for example, a person cannot remember every footstep they take or every sight they encounter. Sensory memory represents the world so that a person gets a sense of time, space and physicality at any moment. But when a sensory memory is noticed or given attention (e.g., a noise, a familiar person, etc.), it then goes into short-term memory, where a person becomes cognisant of the sight or sound; thus, the sensory register is not under cognitive control. If attention is given to a particular sense, then information/sensory data will move to the next store, short-term memory.
Short-term memory is our conscious presence or awareness of the world. Information in STM is fragile, and unless rehearsed, it is still vulnerable to being forgotten due to its limited duration of 18-30 seconds and its finite capacity of 7+ or - 2 items.
If the information in STM is rehearsed and processed deeply enough through maintenance rehearsal - later updated after Craik & Lockhart’s Level of Processing Theory - (LOP) to include elaborative rehearsal - the information can then pass onto the long-term memory store. If information is not (sub) vocalised in the rehearsal loop, it is lost through displacement or trace decay.
LTM is the archive of all memories and is presumed to have unlimited capacity and unlimited duration dependent on the level of processing of the information received. Information retrieval from the long-term store occurs through information passing back through the short-term memory store.
While encoding for long-term memory is mainly semantic and based on meaning, short-term memory is thought to encode acoustically.
A LIST OF RESEARCH STUDIES THAT CAN BE USED FOR EVALUATION
Remember, when you use research studies for A03, you only cite the Findings and Conclusions.
The coding, capacity and duration of each store within the Multi-store model of memory
Jacobs (1887) Capacity/Digit span Experimental cognitive psychology
Miller (1956) Capacity & Chunking Experimental cognitive psychology
Peterson & Peterson (1959) Duration of STM Experimental cognitive psychology
Bahrick et al. (1975) Duration of LTM Field Experiment
Baddeley (1966) Coding of STM & LTM Experimental cognitive psychology
The case study of HM to support and refute the MSM Cognitive neuropsychology
The case study of KF to support and refute the MSM Cognitive neuropsychology
The case of Clive Wearing to support and refute the MSM Cognitive neuropsychology
Murdock (1962) Primacy and Recency and Glanzer also did a similar study called the ‘Serial Position Effect’ supports MSM Experimental cognitive psychology
Craik & Lockhart’s research conflicts with MSM – maintenance rehearsal & elaborative rehearsal Experimental cognitive psychology
Tulving (1972) Types of Long-term memory – episodic, semantic and procedural. Disputes LTM being a unitary store Cognitive neuroscience
Cohen & Squire (1980) differentiate between non-declarative memory and declarative memory. Cognitive neuroscience
Baddeley & Hitch (1975) Dual-task experiments Experimental cognitive psychology
Baddeley & Hitch (1975) Word length effect Experimental cognitive psychology
THE SERIAL POSITION EFFECT (PRIMACY AND RECENCY EFFECT BY GLANZER AND MURDOCK (1962)
Experimental cognitive psychology
The serial position effect (aka primacy and recency effect) is a cognitive phenomenon whereby people tend to remember the first (primacy) and last (recency) items in a series. This provides evidence for the MSM: people tend to remember the first items because they have longer to rehearse the information and may have paid more attention to it, so it is more likely to be transferred to the LTS. They tend to remember the most recent information because it is still in their STS. Information in the middle may be lost because of the limited capacity of the STS. This can be shown in Glanzer and Cunitz’s famous study.
CASE STUDIES ON AMNESIACS
Clive Wearing suffered from herpes encephalitis, which caused extensive damage to his hippocampus and associated structures. This resulted in total amnesia and a profound inability to form new verbal and visual short-term memories (STM). Despite this, his procedural long-term memory (LTM) remained intact; he could still play the piano and conduct music – skills he had acquired before his illness. However, his episodic LTM was severely impaired, leaving him unable to recall personal past experiences, effectively living in a perpetual present.
Henry Molaison (HM): Henry Molaison, often referred to as HM in literature, experienced severe epileptic seizures, which led to a surgical resection to remove his hippocampi and surrounding areas. This operation profoundly impacted his ability to form new verbal STM memories and significantly affected his episodic LTM, preventing the formation of new episodic memories. Interestingly, his procedural LTM was not affected, allowing him to acquire new motor skills, and his semantic LTM remained intact, retaining his general world knowledge.
KF (Full Name Not Publicly Known): KF sustained brain damage from a motorcycle accident, primarily affecting his verbal short-term memory (STM), especially for auditory information. Despite this, KF's memory for visual information within the STM capacity was largely intact, indicating a differentiation in the processing pathways for auditory and visual information. His procedural long-term memory (LTM) remained unaffected, suggesting that the ability to carry out tasks and skills learned before the accident was preserved. While specific impacts on KF's semantic and episodic LTM are not extensively detailed, the preservation of visual STM points to a selective impairment rather than a general memory dysfunction.
The case studies of KF, HM, and Clive Wearing offer intriguing insights into the complexities of human memory and provide both support and challenges to the Multi-Store Model (MSM) of memory.
KF's Case: KF's motorcycle accident caused damage to his left parietal and occipital lobes. His case is peculiar because it shows that while his long-term memory (LTM) was intact, his short-term memory (STM) for verbal information was severely impaired. KF could only recall 1-2 items on immediate memory span tasks, far below the average of 7±2 items that is typically expected. However, his STM for visual information was unscathed, suggesting that STM is not a single, unified store but may have separate subsystems for different types of information. This finding challenges a key aspect of the MSM, which posits a single STM store.
HM's Case: Henry Molaison, known as HM, had his hippocampus removed to alleviate severe epilepsy, leading to significant memory deficits post-surgery. Although HM could maintain verbal information in his STM for about 15 seconds and potentially longer with rehearsal, he could not transfer this information into LTM. He had no issues recalling memories formed before his surgery, yet he couldn't form new long-term memories following the operation. This suggests that the transfer from STM to LTM is a more complex process than the MSM suggests and that there may be different types of STM, given that his ability to learn new motor skills—considered a form of procedural memory—remained intact.
Clive Wearing's Case: Clive Wearing suffered from herpes encephalitis, resulting in extensive damage to his hippocampus and surrounding areas. Like HM, Wearing retained his ability to remember information over short periods and could also learn new tasks, indicating that his procedural memory was preserved. However, he experienced severe anterograde amnesia, unable to form new lasting memories. His case underscores the notion that there are different types of memory stores and processes beyond the simplistic view proposed by the MSM.
These cases collectively indicate that while the MSM provides a foundational framework for understanding memory structures and processes, it may be too limited to account for the observed complexities in human memory functioning. They highlight the likelihood of multiple short-term memory stores and suggest that the encoding and retrieval processes are more intricate than the MSM's linear and singular pathway from STM to LTM.
Clive Wearing was chorus master of London Sinfonietta, BBC radio producer and world music expert. A virus attacked and destroyed his hippocampus and other areas of his cortex. Clive lived in a snapshot of time, constantly believing he had just woken up from a coma. Clive had normal STM but could not lay down new information in LTM. Clive Wearing does not support a singular long-term memory store as he could play the piano and speak (procedural memory) but could not give autobiographical information about his life, e.g., details about his kids.
As case studies also show that LTM and STM have more than one store, make sure you counter-argue this if you use them to support the MSM initially.
CASE STUDY QUESTIONS
A case study was carried out on Peter, whose brain was damaged in a motorcycle accident. Psychologists tested how many numbers he could hold in his short-term memory. They did this by reading him lists of numbers and asking him to recall the numbers immediately in the right order. He could recall a maximum of two items. The psychologists found that his long-term memory was normal.
How was Peter’s short-term memory different from most adults’ after the accident? Two marks
Does this case study support the multi-store model of memory? Explain your answer with four marks.
Identify one ethical issue associated with this case study of Peter. Suggest how psychologists could deal with this ethical issue.
Ethical issue. One mark
How psychologists could deal with this ethical issue. Three marks
SCANS:
Be careful of using scans to support MSM because fMRI, PET, etc, both support and refute MSM.
Brain scans have shown that different brain areas are active when using short-term and long-term memory. Beardsley (1997) found that the prefrontal cortex is active when participants do a task involving their Short-term Memories. Squire et al. (1992) found that the hippocampus is active when participants use their Long-term Memories.
The fact that different brain areas are used depending on which part of the memory is used suggests that the two are separate.
But scans also show that LTM and STM have more than one store, so make sure you counter-argue this if you use them to support the MSM initially.
Brain imaging studies have also been very useful in showing the existence of individual components of working memory. Paulescu et al. (1993) used positron emission tomography (a PET scan) to record brain activity when participants were performing either verbal tasks (using the phonological loop) or visual tasks (using the visuospatial sketch pad). They found that very different brain areas were active during each task. This clearly showed that different brain regions perform verbal and visual tasks, suggesting that the phonological loop and visuospatial sketch pad are physically separate from each other, as the working memory model suggests.
WORKING MEMORY SCAN
TULVING et al (1994): participants performed various memory tasks while their brains were scanned using a PET scanner. and showed that episodic memories and semantic memories had different locations in the cortex.
IS MAINTENANCE REHEARSAL THE ONLY MECHANISM TO GET STM INTO LTM?
Why is the problem of rehearsal a weakness?
The memory of your first kiss doesn’t need to be rehearsed to be remembered, as opposed to trying to remember a phone number. This suggests the MSM model does not explain how certain memories can be transferred into long-term memory without rehearsal.
The levels of processing model (LOP) was proposed as an alternative to the Multistore model. Craik and Lockhart (1972) rejected the idea of separate memory structures put forward by Atkinson and Shiffrin. Their model emphasised memory processes rather than structure, like the MSM model. The LOP model was based on the idea that the strength of a memory trace is determined by how the original information is processed.
Craik and Lockhart didn’t think that STM transferred to LTM by simply maintaining or rehearsing a new memory repetitively via silent or via vocal repetitions.
Craik and Lockhart believed that maintenance rehearsal was a type of memory rehearsal useful in maintaining information in short-term or working memory. Because this usually involves repeating information without thinking about its meaning or connecting it to other information, it is not usually transferred to long-term memory. An example of maintenance rehearsal would be repeating a postcode mentally or aloud until the number is entered into the satellite navigation unit for a journey. The postcode is held in working memory long enough to activate the Satnav but never transferred to long-term memory. The postcode will no longer be remembered an hour or even five minutes after the journey.
Although maintenance has the potential for immediate recall, Craik and Lockhart believed it had little effect on recall in long-term memory. They thought that memory duration depended on the importance of the information that needed to be processed, which would determine which encoding route an individual would use. For example, if the information only needs to be used temporarily, a person will use maintenance rehearsal in working memory. But, if the information needs to be used later, most likely, a person will use elaborative rehearsal. The information is processed deeper in elaborative rehearsal and can move to long-term memory.
Craik and Lockhart proposed that LTM memory encoding involves multiple processes operating at different levels. They believed that maintenance rehearsal was an intermediate level of processing, basically processing a term by sound (not meaning) at the same level at which it entered the brain, e.g., hearing a number or repeating a number. As a result, they believed this process would require little attention.
ELABORATIVE REHEARSAL
Craik and Lockhart reasoned that elaborative rehearsal was the deepest level of memory processing. Elaborative rehearsal is a type of memory rehearsal useful in transferring information into long-term memory. This type of rehearsal is effective because it involves thinking about the meaning of the information and connecting it to other information already stored in memory. For example, in this case, you could remember that 1066 is your best friend Gloria’s house number, but it is also the year of the Battle of Hastings.
It goes much deeper than maintenance rehearsal because of its depth of processing; it requires the learner to engage with new information to create meaningful connections to previously learned things, thus leading to the new information being committed to long-term memory.
HOW TO APPLY ELABORATE REHEARSAL TO YOUR OWN LEARNING - you might as well!
An effective way of encouraging elaborative rehearsal is by engaging with the material in multiple ways. For instance, discussion or study groups allow discrete information to be more personal by attaching stories and creating meaningful connections to things already learned. Elaborative rehearsal strongly supports learning, especially in its attention to meaningful connections across different concepts and pieces of information. More specifically, elaborative rehearsal is extremely beneficial when remembering larger pieces of information such as sentences or other larger chunks.
MULTI-STORE MODEL MARK SCHEME FOR A03
MARK SCHEME MSM EVALUATION (A03)
Possible discussion points:
Useful starting point for memory research, the first model to incorporate three different stores
Evidence supports the three stores' coding, capacity, and duration, e.g., Baddeley, Jacobs, Sperling, Bahrick et al.
Evidence that supports the functional separation of the stores, e.g. Glanzer and Cunitz
Evidence that challenges the unitary nature of STM and LTM, e.g. Shallice and Warrington
Evidence suggests that rehearsal is not the only method of transfer from STM to LTM/ distinction between maintenance and elaborative rehearsal.
Critical comparisons with alternative models, e.g. working memory.
Only credit methodological evaluation of studies if this is used to discuss the strengths/limitations of the model.
Credit other relevant material
PUTTING IT ALL TOGETHER…..
EVALUATION A03 - ESSAY EXAMPLE
The main strength of the MSM comes from support that at least three separate memory stores
exist.
Abundant research supports the multi-store premise that memory is not a singular entity. For example, countless experimental cognitive psychology studies demonstrate that SM, STM and LTM capacities differ vastly. For example, Miller showed that STM had a capacity of 7+ or -2 in most neurotypical people. In contrast, the capacity to keep storing new memories in LTM does not cease even in very old age. For duration studies, Peterson and Peterson showed that STM has a short duration of 18-30 seconds. In contrast, Bahrick showed that high school students remembered their old friends’ photographs many decades after graduation in their LTM. Lastly, Baddeley showed that STM and LTM encode information into the brain differently, with STM encoding mostly acoustically and LTM semantically. However, laboratory experimental studies such as Miller Peterson and Peterson are often criticised for lacking mundane realism because people do not learn nonsense trigrams and meaningless digits in real life. However, other researchers maintain that despite this, such studies have ecological validity as people display similar durations and capacities in their daily lives; this means there is support for the Multi-store of memory.
Research evidence by Glanzer et al. demonstrated support for multi-stores of memory with the serial position effect. Ganzer argued that no matter how many words a person is shown and then asked to recall, items at the beginning of the list are recalled to a greater degree than those in the middle. In contrast, words at the end have a greater recall than the beginning or the middle. Words recalled at the beginning are referred to as the primacy effect, and words remembered at the end of the list are referred to as the recency effect. This supports the MSM because participants remember more words at the beginning of the list since these words are rehearsed and have passed into the LTM (as suggested by the MSM). Words in the middle of the list are also not remembered
because these words are not rehearsed and, therefore, are lost through displacement. Finally, as suggested by the MSM, the words at the end are remembered well because we can hold words in our STM without rehearsal for up to 30 seconds.
The field of cognitive neuropsychology also supports the MSM model with case studies of people with amnesia. For example, Shallice and Warrington (1970) reported the case of KF, who was brain-damaged as a result of a motorcycle accident. His STM was severely impaired; however, his LTM remained intact. This supports the view that it is possible to damage only one store in memory. However, further analysis of such case studies also undermines the MSM as KF demonstrated that his deficit in STM was for verbal information. Still, his STM for visual material was normal. This is a weakness because it shows that it is possible to damage only part of the STM, which refutes a major tenet of MSM by suggesting that there may be more than one type of STM. However, this was a single case study of one individual and may lack external validity to wider generalisation due to possible individual differences.
Other evidence from the discipline of cognitive neuroscience supports the idea of separate short-term and long-term memory stores through PET and fMRI scans that demonstrate different brain activation areas when participants are doing LTM and STM tasks. The prefrontal cortex is seen to relate to short-term memory, while the hippocampus is associated with long-term memory; this is a strength because it provides biological evidence that different parts of the brain process the different types of memory and that the memory stores are distinct as the multi-store model suggests. However, research by Paulescu in 1993 showed that STM may be more complex than Atkinson and Shiffrin had proposed. His research revealed that different brain regions perform verbal and visual tasks. This finding disputes the basic premise of MSM as it suggests STM has more than one store.
This is further compounded by the finding that long-term memory may not be a single store, as Schachter et al. proposed four types of long-term memory stores. These were: Semantic memory is the memory for knowledge; Episodic memory, an individual’s autobiographical story; Procedural memory, which stores actions such as riding a bike or learning to read; and Perceptual-representation systems (PRS), which is memory related to the recognition of specific stimuli. There are multiple areas of support for Schacter’s theory, including Spiers et al., who studied 147 patients with amnesia and found evidence for multiple LTM stores. The case study of Clive Wearing also supports this point, as although Wearing lost his episodic memory, he could still play the piano and play music. These findings suggest Wearing’s procedural memory was still intact and that the multi-store memory model is likely to be an oversimplification of memory structures and processes, e.g., it is extremely likely that there are more than three memory stores and that STM and LTM each have multiple systems.
Another criticism of MSM is from Craik and Lockhart et al., who proposed that rehearsal was not the only way memories are transferred to long-term memory. For example, a first kiss doesn’t need to be rehearsed to be remembered like trying to remember a phone number does. As a result of this appraisal, they proposed that STM memories are encoded into LTM with different levels of processing. For instance, they suggested that memory duration depended on the importance of the information that needed to be processed and that would determine which route of encoding an individual would use, so if a memory only needs to be used temporarily, like learning a number, a person will use maintenance rehearsal But, if the information needs to be used later, a person would probably use elaborative rehearsal. Their research highlighted MSM’s overly simplified explanation of how STM memories are encoded to LTM. Their research resulted in the inclusion of elaborative rehearsal in the revised multi-store memory model.
Further undermining of the MSM comes from suggesting that memory is not a linear system because STM does not always precede LTM. Logie (1999) theorised that STM often relies on LTM because when individuals chunk information in STM, they need to remember the meaning of the items they are chunking so they can be grouped into larger units, e.g., dog, rose, cat, lily, mouse, daisy could get chunked into animal and flower categories of dog, mouse & cat plus lily, daisy & rose. As the meanings of these categories are stored in LTM, Logie proposed that sometimes LTM and STM must be accessed simultaneously. Ruchkin et al. tested this by asking participants to recall real words and pseudo-words. Two large differences were shown in brain scans. If real words and pseudo words involved just STM, then the activity should have been the same for both conditions; however, there was much more activity when real words were processed, indicating the involvement of other brain areas.
It is also possible that individuals are not consciously aware of all their STM memories and that LTM is accessed unknowingly.. For example, if a person speaks to a person with a foreign accent and inadvertently alters their speech to match the foreign speaker’s accent, then they are influenced by what was, until that point, an unconscious (and therefore uncontrollable) aspect of their short-term memory.
It is concluded that STM and LTM might not be separate memory structures, as posited by Atkinson and Shiffrin, but rather work together as memory processes.
Baddeley and Hitch (B&H) criticised the MSM as they said the complexity of STM was ignored. B&H believed that short-term memory was not a passive singular store but had several active processes that could manipulate information. B&H believed there were four separate components in STM which they renamed the Working Memory Model, they were: the phonological loop, the visual spatial sketchpad, the episodic buffer and the central Executive.
The final criticism of MSM concerns the premise that STM is just a place to hold information until it is moved into LTM. In contrast, the concept of the working memory model is that STM is an end in itself as it is a place where information is held until it is put to use, e.g., holding on to what a person has just said to you so you can respond to their comment.
Overall, the Multi-store model was a good forerunner for memory model research , it paved the way for memory to be seen as a system rather than a singular modality, but its premise was rudimentary, and it failed in all aspects to explain the sheer enormity of memory. In the end, memory was discovered to be far more complex than Atkinson and Shiffrin anticipated.
Possible exam questions for the multi-store memory model include:
Outline the multi-store model of memory (6 marks)
Evaluate the multi-store memory model (6 marks AS, 8-10 marks A-level)
Outline and evaluate the multi-store model of memory. 8 marks
Outline and evaluate the multi-store model of memory (12 marks AS, 16 marks A-level)
Outline what psychological research has shown about short-term memory according to the multi-store memory model. 4 marks
Read the item and then answer the questions that follow.
A researcher investigating the multi-store memory model tested short-term memory by reading out loud sequences of numbers that participants had to repeat aloud immediately after the presentation. The first sequence was made up of three numbers: for example, 8, 5, 2. Each participant was tested several times, and each time, the sequence length was increased by adding another number.
Use your knowledge of the multi-store memory model to explain the purpose of this research and the likely outcome.4 marks
After the study was completed, the researcher decided to modify the study by using sequences of letters rather than numbers.
Suggest one 4-letter sequence and one 5-letter sequence that the researcher could use. In the case of each sequence, give a justification for your choice. Use a different justification for each sequence. 4 marks
Psychologists conducted a case study of Patient X, an individual who developed severe amnesia following a car accident. Patient X has difficulty storing new long-term memories, though his short-term memory and memory for events before the accident are unaffected.
Evaluate the use of case studies, like that of Patient X, in psychological research.5 marks
Briefly explain how the experiences of Patient X could be interpreted as supporting the multi-store memory model. 2marks
The question below 3 marks
The multi-store model of memory has been criticised in many ways. The following example illustrates a possible criticism.Some students read through their revision notes lots of times before an examination, but still find it difficult to remember the information. However, the same students can remember the information in a celebrity magazine, even though they read it only once.
12.Explain why this can be used as a criticism of the multi-store model of memory. 4 marks
