Evaluating models of memory | AQA A-Level Psychology Revision
Updated: Sep 9
For 7182 specification, first teach in September 2025
AQA A-Level Psychology | Free Revision Notes
Estimated study time: 45 minutes
These Evaluating models of memory A-Level Psychology revision notes compare the multi-store model and working memory model. Both divide memory into separate parts, but they explain different aspects of remembering. The multi-store model gives a broad account of information moving between sensory, short-term and long-term stores. The working memory model provides a more detailed explanation of active short-term processing. You will examine the evidence supporting each model and consider whether memory can really be represented as a collection of separate boxes or components.
Learning Objectives 🎯
By the end of this revision page, you should be able to:
Compare the structure and purpose of the multi-store and working memory models.
Explain evidence supporting separate short-term and long-term stores.
Explain evidence supporting separate working-memory components.
Evaluate the use of controlled experiments and case studies.
Explain the main limitations of each model.
Assess whether memory can be accurately divided into separate stores or components.
Revision Notes đź“š
Evaluating models of memory A-Level Psychology revision overview
The AQA specification includes two models of memory:
The multi-store model of memory
The working memory model
Both are theoretical representations of memory. They simplify mental processes so that psychologists can describe, investigate and explain them.
However, they have different purposes.
The multi-store model explains the broad movement of information through the memory system.
The working memory model focuses more closely on how information is actively processed during immediate mental tasks.
Neither model should be treated as a literal diagram of physical boxes inside the brain.
Comparing the models
The multi-store model
The multi-store model, or MSM, divides memory into:
The sensory register.
Short-term memory.
Long-term memory.
It proposes that information moves between the stores through:
Attention.
Maintenance rehearsal.
Retrieval.
The model also distinguishes the stores through their coding, capacity and duration.
You can review its complete structure through attention, rehearsal and retrieval.
The working memory model
The working memory model, or WMM, divides active short-term processing into:
The central executive.
The phonological loop.
The visuo-spatial sketchpad.
The episodic buffer.
Each component has a limited capacity and processes information in a different way.
The central executive directs attention and coordinates the other components. The phonological loop processes verbal information, while the visuo-spatial sketchpad processes visual and spatial material. The episodic buffer integrates information and connects working memory with long-term memory.
The components are covered fully in active verbal and visual processing.
Main difference in purpose
The most important difference is the aspect of memory each model attempts to explain.
Multi-store model | Working memory model |
Describes the broad structure of memory | Describes active short-term processing |
Includes the sensory register, STM and LTM | Divides working memory into specialised components |
Explains movement between stores | Explains how different mental tasks are coordinated |
Emphasises attention, rehearsal and retrieval | Emphasises attentional control and specialised processing |
Gives a simple overview of the memory system | Gives a more detailed account of immediate processing |
The WMM does not completely replace the MSM. It develops one area of the earlier model by providing a more detailed account of short-term memory.
Comparing short-term memory
The MSM treats short-term memory as a single store with:
Mainly acoustic coding.
A capacity of approximately 7 ± 2 items.
A duration of approximately 18 to 30 seconds without rehearsal.
The WMM argues that this description is too simple.
It proposes that immediate processing involves several limited-capacity components:
Verbal material is processed by the phonological loop.
Visual and spatial material is processed by the visuo-spatial sketchpad.
Attention is controlled by the central executive.
Information from different sources is integrated by the episodic buffer.
The difference can be summarised as:
MSM: short-term memory is one store.WMM: working memory is a system of specialised components.
Comparing long-term memory
Long-term memory has a central role in the MSM.
The model proposes that information is transferred from STM into LTM through maintenance rehearsal. Stored information is later retrieved from LTM into STM.
The WMM does not attempt to provide a full model of long-term memory.
Instead, it explains how working memory interacts with previously stored information. The episodic buffer allows information retrieved from LTM to be combined with current verbal and visual processing.
It would therefore be inaccurate to criticise the WMM simply because it does not fully describe long-term-memory structure. That is not the model’s main purpose.
Comparing attention
Attention is important in both models.
In the MSM:
Attention selects information from the sensory register.
Attended information enters STM.
In the WMM:
The central executive directs and allocates attention.
It switches attention between tasks.
It coordinates the specialised components.
The WMM therefore gives a more developed account of attentional control.
Comparing capacity
Both models recognise that immediate memory has a limited capacity.
The MSM gives a general estimate of approximately 7 ± 2 items for STM.
The WMM proposes that each component has its own limited capacity. Performance may therefore depend on which component a task uses.
For example, a person may perform a simple verbal and visual task together because they use different components. Two demanding verbal tasks are more likely to interfere because they compete for the phonological loop.
Detailed comparisons of the stores are provided in coding, capacity and duration.
Comparing rehearsal
Maintenance rehearsal is central to the MSM.
The model proposes that repeating information:
Keeps it active in STM.
Helps transfer it into LTM.
Rehearsal also appears in the WMM, but in a more specialised form. The articulatory control process rehearses verbal information within the phonological loop.
The WMM does not claim that all information enters long-term memory simply through verbal repetition. Its focus is how information is temporarily maintained and processed.
Comparing retrieval
The MSM presents retrieval as a movement from LTM back into STM.
The WMM provides more detail about how retrieved information can be used. The episodic buffer combines information from LTM with material currently being processed by the phonological loop and visuo-spatial sketchpad.
For example, when following a familiar route:
Road names may be retained verbally.
Locations may be represented visually and spatially.
Previous knowledge is retrieved from LTM.
The episodic buffer integrates these sources.
Summary comparison table
Feature | Multi-store model | Working memory model |
Main focus | Overall flow of information through memory | Active short-term mental processing |
Main structure | Sensory register, STM and LTM | Central executive, phonological loop, visuo-spatial sketchpad and episodic buffer |
View of STM | One general store | Several specialised components |
Attention | Transfers information from the sensory register to STM | Controlled and allocated by the central executive |
Verbal rehearsal | Maintains information and helps transfer it to LTM | Carried out by the articulatory control process |
Visual processing | Included within a general STM store | Processed by the visuo-spatial sketchpad |
Link with LTM | Rehearsal transfers information to LTM and retrieval returns it to STM | Episodic buffer combines retrieved knowledge with current processing |
Capacity | Gives a general STM estimate | Each component has a limited capacity |
Strength | Clear overview of the entire memory system | Detailed explanation of immediate processing |
Main limitation | Treats STM and LTM too simply | Some components, especially the central executive, remain difficult to define precisely |
Evidence supporting the multi-store model
Different features of the stores
A strength of the MSM is that research suggests the sensory register, STM and LTM have different characteristics.
The stores differ in:
Coding.
Capacity.
Duration.
For example:
Sensory information is coded according to its modality and lasts only very briefly.
STM is mainly acoustic, limited in capacity and brief without rehearsal.
LTM is mainly semantic, potentially unlimited and potentially permanent.
The fact that the stores behave differently supports the idea that they are functionally separate.
Research into coding
Research using different word lists has found that immediate recall is particularly affected when words sound similar.
This supports the claim that STM codes verbal material mainly acoustically.
Following a delay, recall is more affected when words have similar meanings.
This supports the claim that LTM codes verbal information mainly semantically.
These different patterns support the MSM because they suggest that STM and LTM do not represent information in exactly the same way.
However, the findings do not mean that:
STM uses only acoustic coding.
LTM uses only semantic coding.
Both stores can process information in more than one form. The evidence supports a difference in their main coding rather than an absolute division.
Research into capacity
Immediate span tasks show that STM can hold only a limited number of items at once.
Participants are presented with increasingly long sequences and asked to repeat them immediately. The longest sequence recalled provides an estimate of capacity.
This supports the MSM’s claim that STM is a limited store.
Long-term memory, by contrast, appears able to retain a much larger amount of knowledge, experiences and skills.
The difference in capacity supports the distinction between STM and LTM.
Research into duration
Research in which rehearsal is prevented shows that immediate verbal information is rapidly forgotten as the delay before recall increases.
This supports the view that unrehearsed STM has a brief duration.
Research into memories retained over many years suggests that information can remain in LTM for decades.
These findings support the MSM because the proposed stores have very different durations.
The serial-position effect
Evidence for separate STM and LTM comes from the serial-position effect.
When people recall a list of words, they often remember:
Words at the beginning of the list.
Words at the end of the list.
Recall of words in the middle is usually poorer.
The improved recall of early words is known as the primacy effect. The MSM explains this by proposing that early items receive more rehearsal and enter LTM.
The improved recall of final words is known as the recency effect. The MSM explains this by proposing that the final items remain available in STM at the point of immediate recall.
Delayed recall and the recency effect
When recall is delayed by an activity that prevents rehearsal, the recency effect is reduced.
The final words are no longer available in STM because its limited duration has been exceeded.
The primacy effect may remain because earlier words have already entered LTM.
This supports the distinction between:
A brief STM responsible for recent items.
A more durable LTM responsible for early, rehearsed items.
The strength of this evidence is that one recall task produces two different patterns that can be explained by separate stores.
Selective damage
Evidence from people with memory impairments can also support separate stores.
Some individuals can hold information temporarily but have great difficulty creating durable new memories. This suggests that immediate memory and long-term storage can be affected differently.
If STM and LTM were one undivided system, it would be difficult to explain why one could remain more functional than the other.
However, evidence from unusual cases must be interpreted carefully because the pattern of damage may be complex and may not represent memory in everyone.
Limitations of the multi-store model
Short-term memory may not be one store
A major limitation is that the MSM treats STM as a single unit.
The WMM provides evidence that immediate memory contains specialised systems for:
Verbal information.
Visual information.
Spatial information.
Attentional control.
Integrating different sources.
A person may perform a verbal and visual task together more successfully than two verbal tasks. This is difficult to explain if all short-term processing occurs within one undivided store.
The WMM therefore suggests that the MSM’s account of STM is too simple.
Long-term memory may also be oversimplified
The MSM treats LTM as one general store.
However, people may remember some forms of information while having difficulty with others. This suggests that long-term storage may be more complex than a single box.
For the present specification, the important conclusion is that treating all long-term information as one identical type of memory may be an oversimplification.
Rehearsal may not be sufficient for long-term learning
The MSM gives maintenance rehearsal a central role in transferring information into LTM.
However, simple repetition does not always produce durable understanding.
A student might repeat a definition many times without understanding its meaning and then forget it soon afterwards. By contrast, meaningful information may sometimes be remembered even when it has not been rehearsed repeatedly.
This suggests that:
The nature of processing may matter.
Meaning may be more important than repetition alone.
Information does not always move into LTM through one fixed route.
Maintenance rehearsal can help learning, but it may not provide a complete explanation of
long-term storage.
The flow of information may be too linear
The MSM presents memory as a sequence:
Sensory register → STM → LTM
This is clear and testable, but everyday memory may not always operate in such a simple order.
Existing knowledge in LTM can influence:
What a person notices.
How new information is interpreted.
What receives attention.
How information is organised in immediate memory.
The relationship between working memory and LTM may therefore be more interactive than the MSM’s one-directional flow initially suggests.
Artificial research tasks
Much of the evidence supporting separate stores comes from tasks involving:
Word lists.
Digits.
Trigrams.
Immediate recall.
Brief laboratory procedures.
These methods allow high levels of control. Researchers can standardise the material, delays and instructions.
However, remembering unrelated words may differ from remembering:
A meaningful conversation.
Course content.
A personal experience.
A route through a familiar place.
Instructions used during a practical task.
This may limit the ecological validity of the findings.
The criticism does not make the research worthless. Controlled tasks help isolate coding, capacity and duration. The limitation is that the conclusions may not fully explain complex everyday remembering.
Evidence supporting the working memory model
Dual-task research
A major strength of the WMM is evidence from dual-task studies.
A dual-task study asks participants to perform two activities at the same time.
The model predicts that:
Two tasks using different components should produce relatively little interference.
Two tasks using the same component should produce greater interference.
Two unfamiliar tasks may also compete for central executive attention.
For example, a verbal task may be completed alongside a simple visual task more successfully than alongside another demanding verbal task.
This supports the idea that working memory contains specialised components rather than one general short-term store.
Why dual-task evidence is important
Dual-task research does more than show that multitasking can be difficult.
The pattern of interference matters.
If all tasks used one single STM store, any two equally demanding tasks should interfere in broadly similar ways.
Instead, interference depends partly on the type of information being processed. This supports the separation between:
The phonological loop.
The visuo-spatial sketchpad.
Evidence from the case of KF
The case of KF provides further support for separate components.
KF experienced impaired verbal and auditory short-term processing while visual-spatial processing was relatively intact.
This supports the WMM because it suggests that:
Verbal and visual short-term memory can be affected separately.
The phonological loop and visuo-spatial sketchpad are functionally distinct.
Immediate memory is unlikely to be one unitary store.
It is important not to reverse the pattern. KF’s visual-spatial processing was more intact than his verbal or auditory processing.
Strengths of case evidence
Evidence from a person with selective impairment can reveal distinctions that may be difficult to observe in participants without memory difficulties.
If one ability is damaged while another remains relatively intact, psychologists can infer that the two abilities may depend on different systems.
This is particularly valuable for evaluating whether short-term memory has separate components.
Limitations of case evidence
A case study involves an unusual individual.
This creates several limitations:
The exact brain damage may be unique.
The person may have had other difficulties.
Previous abilities are not always known with certainty.
Findings may not generalise to everyone.
The damage may affect several interacting systems.
The case of KF supports the distinction between verbal and visual-spatial processing, but it cannot prove that the WMM is complete or correct in every detail.
Practical explanatory value
The WMM can explain everyday differences in performance.
For example, it explains why someone might:
Listen to spoken instructions while copying a simple diagram.
Struggle to listen to a conversation while reading complex text.
Find two visual-spatial tasks difficult to perform together.
Find multitasking harder when both activities are unfamiliar.
This gives the model practical value because it connects the structure of working memory with recognisable behaviour.
Different components make testable predictions
The model makes specific predictions.
For example:
Verbal suppression should disrupt verbal rehearsal.
A visual-spatial task should interfere with another visual-spatial activity.
Familiar tasks should place fewer demands on the central executive.
Combining current information with previous knowledge should involve the episodic buffer.
These predictions can be investigated experimentally, making the model scientifically useful.
Limitations of the working memory model
The central executive is vague
A major limitation is that the central executive is not described in much detail.
It is said to:
Direct attention.
Divide resources.
Switch between tasks.
Coordinate other components.
Monitor performance.
This is a wide range of functions.
Describing it as the system that controls everything not explained by another component risks making it too broad.
Important questions remain:
Is it one component or several processes?
How does it choose which task receives attention?
How exactly are resources allocated?
How can its capacity be measured?
The central executive may therefore be less clearly defined than the phonological loop or visuo-spatial sketchpad.
The components may be more complex
The WMM improves on a single STM store by dividing working memory into components.
However, each component may itself contain several processes.
For example:
Verbal processing includes storage and rehearsal.
Visual processing differs from spatial processing.
Attentional control may involve several different abilities.
Integration may operate differently across tasks.
The model may therefore replace a small number of broad stores with a larger number of still-simplified components.
The episodic buffer is difficult to isolate
The episodic buffer explains how information from different sources is integrated.
This is useful because many tasks require combined verbal, visual and long-term information.
However, it can be difficult to determine which behaviour is specifically caused by the episodic buffer rather than:
The central executive.
The phonological loop.
The visuo-spatial sketchpad.
Long-term-memory knowledge.
A component that is difficult to isolate may also be difficult to test independently.
Artificial dual-task procedures
Dual-task studies often use carefully controlled activities.
This is a strength because task demands can be manipulated and compared.
However, laboratory tasks may not reflect everyday multitasking.
Real activities often involve:
Meaningful goals.
Practice and expertise.
Emotional demands.
Continuous feedback.
Changing levels of difficulty.
More than two sources of information.
Laboratory findings support specialised working-memory components, but they may not capture the full complexity of working memory outside controlled settings.
The WMM is not a complete model of memory
The WMM focuses on active short-term processing.
It does not fully explain:
The sensory register.
How durable long-term memories are formed.
The complete structure of LTM.
Every reason why information is forgotten.
This limits its scope, but it is not necessarily a weakness in the model itself. A model should be judged primarily against the processes it was designed to explain.
The WMM can therefore be detailed but narrow, while the MSM is broad but simpler.
Are memory stores really separate?
Functional separation
Evidence suggests that parts of memory perform different functions.
For example:
Immediate verbal and visual tasks can show different interference patterns.
STM and LTM have different coding, capacities and durations.
Selective impairments can affect one form of memory more than another.
This supports some degree of functional separation.
However, functional separation does not mean the systems operate independently at all times.
The systems interact
During most everyday tasks, several memory processes interact.
When reading a paragraph:
Visual information must be processed.
Words may be represented verbally.
Earlier sentences must remain available.
Meanings are retrieved from LTM.
Attention is directed towards relevant information.
The ideas are combined into a coherent understanding.
It can be useful to separate these functions for research, but actual remembering usually depends on their cooperation.
Boxes may encourage oversimplification
Diagrams can make memory models easier to understand, but they can also create a misleading impression.
The boxes may suggest that:
Each component has a clear boundary.
Information belongs to only one store at a time.
Each store performs one fixed function.
Processes occur in a strict sequence.
The stores correspond to single physical locations.
The models are better understood as simplified descriptions of psychological functions.
Components may overlap
A mental task rarely uses only one component.
Reading a map may involve:
The visuo-spatial sketchpad for locations.
The phonological loop for road names.
The central executive for selecting a route.
The episodic buffer for integrating the information.
LTM for recognising familiar places.
This makes it difficult to isolate a component in a completely pure form.
Performance on a task may reflect several processes rather than one store alone.
Research tasks may create apparent separation
Researchers design tasks to isolate one memory function.
For example, they may use:
Similar-sounding word lists to test acoustic coding.
Visual patterns to test spatial processing.
Counting backwards to prevent rehearsal.
Two simultaneous tasks to test interference.
These methods are valuable, but the task design may exaggerate distinctions that are less clear in ordinary behaviour.
A finding that two laboratory tasks differ does not necessarily prove that memory contains perfectly separate, independent boxes.
Individual differences
Memory capacity and processing vary between individuals.
Performance may be affected by:
Age.
Experience.
Familiarity with the material.
Practice.
The use of strategies.
The meaningfulness of the task.
The ability to divide attention.
A fixed diagram may not represent these individual differences easily.
Models are useful simplifications
The fact that models simplify memory does not make them useless.
A useful model should:
Organise evidence.
Define key processes.
Generate testable predictions.
Explain patterns of behaviour.
Allow comparison with alternative accounts.
Be revised when new evidence becomes available.
The MSM and WMM both meet several of these requirements.
The more balanced conclusion is:
Memory models are useful representations of different functions, but their stores and components should not be treated as perfectly independent or literal structures.
Which model is more useful?
There is no single answer because the models serve different purposes.
When the MSM is useful
The MSM is particularly useful for explaining:
The general route from sensory input to long-term storage.
Differences in coding, capacity and duration.
The role of attention.
Maintenance rehearsal.
Retrieval.
Primacy and recency effects.
When the WMM is useful
The WMM is particularly useful for explaining:
Active short-term processing.
Divided attention.
Verbal rehearsal.
Visual and spatial processing.
Dual-task performance.
Integration of current information with LTM.
A balanced judgement
The WMM provides a more convincing account of the complexity of immediate memory because it is supported by dual-task and selective-impairment evidence.
However, it has a narrower focus and some components remain difficult to define.
The MSM offers a clear framework for the wider memory system, but its description of STM as one store and its emphasis on a simple linear flow are limited.
The models are therefore best viewed as:
Different levels of explanation.
Useful for different questions.
Simplified rather than complete accounts.
Evaluating evidence effectively
An evaluation paragraph should do more than describe a study.
Use this structure:
State the evidence.
Explain the relevant finding.
Link the finding to a feature of the model.
Reach a clear judgement.
Weak evaluation
KF had brain damage, which supports the working memory model.
This does not explain what the evidence demonstrates.
Stronger evaluation
KF had impaired verbal short-term processing while visual-spatial processing was relatively intact. This selective impairment supports the WMM because it suggests that verbal and visual material depend on separate components rather than one unitary STM store.
Weak methodological criticism
The research was artificial.
This lacks explanation.
Stronger methodological criticism
Dual-task studies often use artificial verbal and visual activities. These may not represent everyday tasks in which experience, meaning and feedback influence performance, so the evidence may provide an incomplete picture of working memory outside the laboratory.
Planning a comparison answer
When a question asks you to compare the models, organise the answer by shared features rather than writing everything about one model and then everything about the other.
A useful structure is:
Compare their overall aims.
Compare their structures.
Compare their accounts of STM.
Compare their treatment of attention.
Compare rehearsal and retrieval.
Compare their evidence.
Compare their limitations.
Reach a balanced conclusion.
Use comparative language throughout:
“Whereas”
“In contrast”
“Both models”
“Unlike the MSM”
“The WMM provides more detail”
“The MSM has a broader scope”
Hints from the Examiner Reports đź’ˇ
Examiner hint:Â Do not confuse the two models. The sensory register, STM and LTM belong to the multi-store model. The central executive and specialised components belong to the working memory model.
Examiner hint:Â Include relevant characteristics when describing a model. Stronger answers refer to coding, capacity and duration in the MSM and coding and capacity in the WMM.
Examiner hint:Â When using the case of KF, state the selective pattern accurately. Verbal or auditory short-term processing was impaired, while visual-spatial processing was relatively intact.
Examiner hint:Â Do not confuse the episodic buffer with episodic long-term memory. The episodic buffer temporarily integrates information and links working memory with LTM.
Examiner hint:Â Do not criticise the WMM simply for failing to explain the full structure of LTM. It was designed primarily as a model of active short-term processing.
Examiner hint: Evaluation must be linked to the model. Describing a study’s artificial procedure is not enough unless you explain how this affects the strength of the evidence.
Examiner hint:Â Use comparative wording when the question asks you to compare. Two separate descriptions may demonstrate knowledge but fail to make the similarities and differences explicit.
Examiner hint:Â Avoid overloading an answer with procedural details. Select the finding that supports or challenges the relevant feature of the model.
Common Mistakes ⚠️
Mistake: Saying that the WMM replaces the whole multi-store model
Why this is incorrect:
The WMM focuses mainly on active short-term processing. It does not provide a complete alternative account of the sensory register and LTM.
How to improve:
State that the WMM develops the MSM’s account of short-term memory.
Mistake: Treating the models as identical
Why this is incorrect:
The MSM describes broad information flow, while the WMM explains specialised immediate processing.
How to improve:
Compare their aims, structures and treatment of STM.
Mistake: Writing two separate descriptions in a comparison question
Why this is incorrect:
A comparison requires explicit similarities and differences.
How to improve:
Organise the answer by features and use words such as “whereas”, “both” and “in contrast”.
Mistake: Saying that serial-position evidence proves the MSM
Why this is incorrect:
The findings are consistent with separate short-term and long-term processes, but one pattern of evidence cannot prove that every part of the model is correct.
How to improve:
State that the evidence supports the distinction between STM and LTM.
Mistake: Describing KF’s impairments the wrong way round
Why this is incorrect:
KF had greater difficulty with verbal and auditory short-term processing, while visual-spatial processing was relatively intact.
How to improve:
Link the correct pattern to separate working-memory components.
Mistake: Using a case study without evaluating it
Why this is incorrect:
Selective impairment can provide valuable evidence, but unusual cases may not generalise to everyone.
How to improve:
Explain both the insight and the limitation.
Mistake: Claiming that different components operate independently
Why this is incorrect:
The central executive coordinates them, and the episodic buffer combines their information. Most everyday tasks use several components.
How to improve:
Describe the components as specialised but interacting.
Mistake: Treating the model diagrams as literal brain structures
Why this is incorrect:
The boxes represent psychological functions and processes rather than physical containers.
How to improve:
Refer to functional systems or theoretical components.
Mistake: Giving generic criticism of laboratory research
Why this is incorrect:
A broad statement about artificiality does not evaluate the model.
How to improve:
Explain why artificial word lists or dual tasks may not represent meaningful everyday remembering.
Mistake: Saying that rehearsal is the only way information reaches LTM
Why this is incorrect:
The MSM emphasises rehearsal, but meaningful information may be retained without extensive repetition, and repetition does not always produce lasting understanding.
How to improve:
Evaluate maintenance rehearsal as useful but incomplete.
Mistake: Criticising the WMM for not describing all of LTM
Why this is incorrect:
The model was not designed to provide a complete explanation of long-term storage.
How to improve:
Evaluate how successfully it explains working memory.
Exam-Style Questions ✍️
Question 1
Which one of the following is a component of the working memory model but not a store in the multi-store model?
A. Sensory register
B. Long-term memory
C. Phonological loop
D. Short-term memory
[1 mark]
reveal mark scheme and indicative content
Answer: C. Phonological loop
Question 2
Explain one difference between the multi-store model and the working memory model.
[3 marks]
reveal mark scheme and indicative content
Mark scheme
AO1 = 3
Award marks for a clear comparison:
1 mark:Â Accurate feature of the multi-store model.
1 mark:Â Contrasting feature of the working memory model.
1 mark:Â Clear elaboration of the difference.
Indicative content
The multi-store model treats STM as a single store.
The working memory model divides short-term processing into several components.
The multi-store model describes memory as relatively passive and linear, whereas the working memory model emphasises active processing.
Example answer
The multi-store model treats short-term memory as one unitary store, whereas the working memory model divides short-term processing into components such as the phonological loop and visuo-spatial sketchpad. The working memory model therefore gives a more detailed account of short-term processing.
Question 3
Explain how the serial-position effect supports the distinction between short-term and long-term memory.
[4 marks]
reveal mark scheme and indicative content
Mark scheme
AO3 = 4
Award 1 mark each for:
Better recall of the first words is the primacy effect.
Early words can be rehearsed and transferred to LTM.
Better recall of the final words is the recency effect.
Final words are still available in STMÂ during immediate recall, supporting separate stores.
AQA's 2023 mark scheme explains primacy in terms of transfer to LTM and recency in terms of the final items remaining in STM.
Example answer
The serial-position effect shows better recall of words at the beginning and end of a list. Early words produce a primacy effect because they receive more rehearsal and enter long-term memory. Final words produce a recency effect because they are still present in short-term memory. This supports the idea that STM and LTM are separate stores.
Question 4
Outline one limitation of the multi-store model’s account of short-term memory.
[2 marks]
reveal mark scheme and indicative content
Mark scheme
AO3 = 2
2 marks:Â Clear limitation with elaboration.
1 mark:Â Limited or undeveloped limitation.
Indicative content
The model may oversimplify STM by treating it as a single unitary store.
Evidence from KF and dual-task research suggests separate verbal and visual short-term components.
Example answer
One limitation is that the multi-store model treats STM as a single store. Evidence that verbal and visual short-term memory can be affected separately suggests that STM contains several components instead.
Question 5
A participant can repeat a short list of spoken words but cannot perform a verbal reasoning task while repeating an unrelated word. The same participant can perform the reasoning task while copying a simple visual pattern.
Use the working memory model to explain this pattern.
[4 marks]
reveal mark scheme and indicative content
Mark scheme
AO2 = 4
Award 1 mark each for:
Verbal material is processed by the phonological loop.
Repeating an unrelated word occupies the same limited-capacity verbal component.
This interferes with the verbal reasoning task.
Copying a visual pattern uses the visuo-spatial sketchpad, so there is less competition with the verbal task.
Example answer
The verbal reasoning task and repeating an unrelated word both require the phonological loop, so they compete for its limited capacity and performance is impaired. Copying a visual pattern mainly uses the visuo-spatial sketchpad instead. Because the two tasks use different components, the participant can perform them together more successfully.
Question 6
Explain how the case of KF provides support for the working memory model.
[4 marks]
reveal mark scheme and indicative content
Mark scheme
AO3 = 4
Award 1 mark each for:
KF had impaired short-term memory for verbal/acoustic information.
His short-term memory for visual information was relatively intact.
This suggests verbal and visual information are processed separately.
This supports separate components such as the phonological loop and visuo-spatial sketchpad, rather than one unitary STM.
AQA specifically identifies KF as supporting separate visual and verbal stores in STM, and warns that his visual processing was relatively intact while verbal processing was impaired.
Example answer
KF had poor short-term memory for verbal information but relatively intact short-term memory for visual information. If STM were one single store, both types of information should have been similarly affected. His case therefore supports the working memory model's distinction between the phonological loop and visuo-spatial sketchpad.
Question 7
A student says:
“The working memory model must be wrong because it does not explain the different kinds of information stored in long-term memory.”
Explain why this is not an effective criticism of the working memory model.
[4 marks]
reveal mark scheme and indicative content
Mark scheme
AO3 = 4
Award 1 mark each for:
The working memory model was designed primarily to explain short-term/working memory.
It was not intended to provide a detailed model of the different forms of LTM.
Therefore, failure to describe LTM structure does not directly challenge what the model claims to explain.
The episodic buffer nevertheless provides a link between working memory and LTM.
AQA's 2022 examiner report explicitly states that criticising the working memory model for not explaining the structure of LTM is not creditworthy because the model was not designed to do this.
Example answer
This is not an effective criticism because the working memory model was designed to explain short-term processing, not the detailed structure of long-term memory. Therefore, failing to distinguish different types of LTM does not contradict the model's purpose. The episodic buffer also provides a link between working memory and LTM.
Question 8
Two researchers investigate memory.
Researcher A gives participants lists of unrelated words and measures immediate and delayed recall.
Researcher B asks participants to complete verbal and visual tasks separately and simultaneously.
Explain how each investigation could provide evidence relevant to a different model of memory.
[6 marks]
reveal mark scheme and indicative content
Mark scheme
AO2 = 6
Award up to 3 marks for each researcher.
Researcher A
Relevant to the multi-store model.
Immediate recall may produce recency because final items remain in STM.
Delayed recall can reduce recency while primacy remains, supporting separate STM and LTM.
Researcher B
Relevant to the working memory model.
Verbal and visual tasks can be compared in single-task and dual-task conditions.
Less interference between tasks using different components would support separate phonological and visuo-spatial systems.
Example answer
Researcher A could provide evidence for the multi-store model. Immediate recall may show both primacy and recency, whereas delayed recall may reduce the recency effect because information has left STM, supporting separate STM and LTM stores. Researcher B could test the working memory model. If people can perform a verbal and visual task together better than two tasks using the same type of processing, this would support separate components within working memory.
Question 9
Compare the multi-store model and the working memory model.
[8 marks]
reveal mark scheme and indicative content
Mark scheme
AO1/AO3 = 8
7–8 marks: Comparison is clear, accurate and detailed. Several meaningful similarities and differences are explained using appropriate terminology.
5–6 marks: Mostly clear comparison with relevant similarities and differences, although some detail may be lacking.
3–4 marks: Some relevant knowledge but comparison is limited or uneven.
1–2 marks: Very limited comparison, perhaps consisting mainly of separate descriptions.
0 marks:Â No relevant content.
Indicative content
Similarities
Both are cognitive models of memory.
Both propose separate components/stores rather than memory being a single undifferentiated system.
Both recognise limitations in short-term processing.
Differences
The multi-store model includes the sensory register, STM and LTM, whereas the working memory model focuses mainly on short-term processing.
The multi-store model treats STM as unitary, whereas the working memory model divides it into the central executive, phonological loop, visuo-spatial sketchpad and episodic buffer.
The multi-store model describes movement through stores using processes such as attention, rehearsal and retrieval.
The working memory model focuses on active processing, divided attention and dual-task performance.
The working memory model provides a more detailed explanation of why verbal and visual tasks can sometimes be performed simultaneously.
Evidence such as KFÂ challenges a single STM store and supports the working memory model.
AQA describes the working memory model as an active account of STM and credits KF and dual-task evidence as support for separate components.
Example answer
Both models argue that memory consists of separate parts rather than being a single system. However, the multi-store model proposes three main stores: the sensory register, STM and LTM, and explains how information moves between them through attention, rehearsal and retrieval. In contrast, the working memory model focuses mainly on short-term processing and divides it into the central executive, phonological loop, visuo-spatial sketchpad and episodic buffer.
The multi-store model therefore treats STM as relatively unitary and passive, whereas the working memory model sees it as an active system that can process different types of information simultaneously. Evidence from KF supports the working memory model because his verbal STM was impaired while visual STM was relatively intact, suggesting separate short-term components rather than the single STM proposed by the multi-store model.
Question 10
Discuss the strengths and limitations of describing memory as a collection of separate stores or components. Refer to the multi-store model and working memory model in your answer.
[16 marks]
reveal mark scheme and indicative content
Comments