Memory — AQA GCSE Psychology
Memory explained
Memory covers 10 specification points, including Encoding (input) storage and retrieval (output) and Episodic memory, semantic memory and….
Read the full explanation
Open a point below for its explanation and revision guidance.
Processes of memory: encoding (input) storage and retrieval (output)
Memory is best understood as a flow of information through three processes. Encoding is the input stage: information is converted into a form the memory system can hold, for example turning a spoken word into its sound pattern or a picture into a visual code. Storage is keeping that encoded information available over time, as when a phone number is held in mind or a fact is filed away. Retrieval is the output stage: stored information is located and brought back into awareness, for example recalling a name when you see a face. A useful check is that encoding happens before storage, and storage happens before retrieval; if encoding is weak, later retrieval fails even when storage was attempted. Exam questions often ask you to identify which process a scenario describes, so match the verb in the scenario to input, keeping or output.
Different types of memory: episodic memory, semantic memory and procedural memory.
Long-term memory is not one single store; it holds different types of knowledge. Episodic memory records personal events tied to a time and place, such as remembering your first day at a new school. Semantic memory holds general facts and meanings that are not tied to when you learned them, such as knowing that Paris is the capital of France. Procedural memory stores the skills needed to perform actions, such as riding a bike or tying shoelaces, and these skills are often hard to describe in words. A reliable way to classify a scenario is to ask three questions: is it a personal event with a time and place (episodic), a fact or meaning (semantic), or a how-to skill (procedural)? Some examples overlap, so choose the type that best matches the main content of the scenario.
How memories are encoded and stored.
Encoding is the process of converting information into a form the memory system can hold; storage is keeping that encoded information available over time. Learners should distinguish the two processes and link them to the multi-store model: sensory memory holds brief raw impressions, short-term memory holds a limited amount for a short time, and long-term memory holds a large amount for a long time. Encoding can be visual (as when holding an image of a shape just seen), acoustic (as when repeating a phone number in your head) or semantic (as when storing the meaning of a fact). A useful method is to take one example, such as hearing a new word, and trace it: it is encoded acoustically or semantically, then stored, and later retrieved.
Structures of memory The multi-store model of memory: sensory, short term and long term.
The multi-store model describes memory as a set of separate stores through which information passes in sequence. Information from the environment first enters sensory memory, which holds a brief impression for a very short time. If attention is paid, information transfers to short-term memory, which has a limited capacity and duration. With rehearsal, information can pass into long-term memory, which has a large capacity and long duration. Each store differs in capacity, duration and encoding. A useful method is to sketch the model as boxes joined by arrows labelled attention and rehearsal, then test yourself on what each store holds and how information moves between them. MCQ items ask you to identify a store from its described features or to recognise the transfer process, so match the features in the stem to the correct store.
Features of each store: coding, capacity, duration.
The multi-store model treats memory as separate stores, and each store is described by three features. Coding is the form in which information is held: sensory memory codes modality-specifically (for example iconic visual or echoic auditory), short-term memory tends to code acoustically, and long-term memory tends to code semantically. Capacity is how much can be held: sensory stores hold a great deal briefly, short-term memory is limited (often cited as about 7 ± 2 items), and long-term memory is effectively unlimited. Duration is how long information lasts: sensory memory lasts a fraction of a second to a few seconds, short-term memory around 18–30 seconds without rehearsal, and long-term memory potentially a lifetime. Learn each store against all three features and be ready to match a described feature to the correct store.
Primacy and recency effects in recall: the effects of serial position.
When people recall a list in any order (free recall), the position of each item affects whether it is remembered. Items at the start of the list are recalled well — the primacy effect — because they receive more rehearsal and are transferred to long-term memory. Items at the end are also recalled well — the recency effect — because they are still in short-term memory at the time of recall. Items in the middle are recalled least well. Plotting percentage recall against serial position produces a U-shaped curve. If recall is delayed or a distracting task is given before recall, the recency effect largely disappears because the last items are displaced from short-term memory, while the primacy effect remains. Learn to explain both effects and the shape of the curve.
Murdock’s serial position curve study.
Murdock investigated free recall of word lists to test whether position in a list affects memory. Participants heard lists of words at different lengths and recalled them in any order. When recall was plotted by position, a serial position curve appeared: words at the start (primacy) and end (recency) of the list were recalled better than those in the middle. Murdock explained primacy as words having been rehearsed into long-term memory, and recency as words still in short-term memory. The middle words suffered because they had been displaced from short-term memory before entering long-term memory. This supports the multi-store model's distinction between short-term and long-term memory. For an MCQ, identify the pattern, the explanation of primacy and recency, and the study's design features.
Memory as an active process The Theory of Reconstructive Memory, including the concept of ‘effort after meaning’.
The Theory of Reconstructive Memory, proposed by Bartlett, treats remembering as an active process rather than a passive replay of stored information. When we recall an event we use schemas, mental frameworks built from past experience, to fill gaps and make sense of what we remember. This means memories can be distorted, simplified or changed to fit existing expectations. The concept of 'effort after meaning' describes the mental work people do to interpret and give meaning to new material, drawing on their own knowledge and culture. Bartlett's War of the Ghosts study showed participants reshaping an unfamiliar story to fit their own cultural schemas. For an MCQ, distinguish active reconstruction from passive storage, and recognise effort after meaning as the attempt to make sense of information.
Bartlett’s War of the Ghosts study.
Bartlett investigated reconstructive memory using a Native American folk story, 'The War of the Ghosts', which was unfamiliar to his British participants. He asked them to read the story and then reproduce it repeatedly over time. Participants changed the story: they shortened it, made it more coherent, and altered unfamiliar details so they fitted their own cultural expectations, a process Bartlett called effort after meaning. He argued that memory is not a passive store but an active reconstruction shaped by schemas. In an MCQ, expect questions on the method, the findings and the conclusion that recall is reconstructive and influenced by cultural schemas.
Factors affecting the accuracy of memory, including interference, context and false memories.
Accuracy of memory is affected by several factors. Interference occurs when one memory disrupts another; proactive interference is when an older memory disrupts a newer one, and retroactive interference is when a newer memory disrupts an older one. Context affects accuracy because recall is better when the environment at recall matches the environment at learning, known as context-dependent forgetting when they differ. False memories are inaccurate recollections of events that did not happen or happened differently, and leading questions or post-event information can create them, as shown in Loftus and Palmer's research on eyewitness testimony. In an MCQ, expect questions distinguishing these factors and identifying examples of each.
Watch and listen
Listen: Memory audio lesson
Read the full guideYour focus
- Define encoding, storage and retrieval accurately.
- Identify which memory process a given scenario illustrates.
- Explain why the order of the three processes matters for successful recall.
Show all 30 objectives
- Describe episodic, semantic and procedural memory.
- Classify given examples into the correct memory type.
- Distinguish between factual knowledge and skilled action in memory examples.
- Define encoding and storage and state how they differ.
- Identify whether a given scenario illustrates encoding, storage or retrieval.
- Recognise visual, acoustic and semantic encoding in examples.
- Describe the sensory, short-term and long-term stores of the multi-store model.
- Identify which store is described by given capacity, duration or encoding features.
- Explain how attention and rehearsal move information between stores.
- Identify the coding, capacity and duration of sensory, short-term and long-term memory stores.
- Match a described feature to the correct memory store.
- Explain how the three features distinguish one store from another.
- Describe the primacy and recency effects in free recall.
- Explain each effect in terms of short-term and long-term memory.
- Predict how a delay before recall changes the serial position curve.
- Describe the pattern of the serial position curve found by Murdock.
- Explain primacy and recency in terms of short-term and long-term memory.
- Identify the design features of Murdock's study, including free recall of word lists.
- Explain reconstructive memory as an active process guided by schemas.
- Define and apply the concept of 'effort after meaning'.
- Describe how reconstruction can distort memories, using Bartlett's research as evidence.
- Describe the method used by Bartlett in the War of the Ghosts study.
- Explain the findings in terms of schema-driven distortion.
- State Bartlett's conclusion that memory is an active reconstruction.
- Distinguish proactive and retroactive interference with examples.
- Explain how context affects the accuracy of recall.
- Describe how false memories arise and affect accuracy.
Memory flashcards
Card 1 of 10
Encoding
Memory exam tips
Marking Points
- Encoding is the input process in which information is changed into a form that can be stored, such as acoustic or visual coding.
- Storage is the process of holding encoded information over time so that it remains available.
- Retrieval is the output process of locating and bringing stored information back into use.
- The order is encoding, then storage, then retrieval, and a weakness at an earlier stage limits later stages.
- In a scenario, the process is identified by what is happening: taking information in, keeping it, or bringing it back.
- Episodic memory stores personal experiences and events linked to a particular time and place.
- Semantic memory stores general knowledge, facts and meanings that are not tied to a specific personal event.
- Procedural memory stores the knowledge of how to carry out skills and actions.
- Classifying a scenario depends on whether it describes a personal event, a fact, or a skill.
- Procedural skills are often difficult to explain verbally even when they can be performed.
- Encoding is the conversion of information into a form that can be held in memory; storage is the retention of that encoded information over time.
- Sensory memory, short-term memory and long-term memory differ in capacity and duration, and encoding and storage operate within each store.
- Encoding may be described as visual, acoustic or semantic, and a scenario should be matched to the type of code it illustrates.
- Storage is not the same as retrieval: retrieval is accessing stored information, whereas storage is keeping it available.
- A correct answer identifies the process or code named in the stem and rejects distractors that describe a different process.
- The multi-store model proposes separate sensory, short-term and long-term stores linked in sequence.
- Sensory memory holds brief impressions for a very short duration and has a large capacity.
- Short-term memory has limited capacity and duration and relies on rehearsal to retain information.
- Long-term memory has a large capacity and long duration and is reached through rehearsal from short-term memory.
- Attention moves information from sensory memory to short-term memory, and rehearsal moves it from short-term to long-term memory.
- Coding is the format in which information is stored; sensory memory is modality-specific, short-term memory is mainly acoustic and long-term memory mainly semantic.
- Capacity is the amount a store can hold; short-term memory is limited to roughly 7 ± 2 items, whereas long-term memory is considered effectively unlimited.
- Duration is how long information persists; sensory memory lasts a fraction of a second to a few seconds, short-term memory about 18–30 seconds without rehearsal, and long-term memory potentially a lifetime.
- Each store must be described using all three features, so a correct answer links a named store to its coding, capacity and duration rather than giving one feature alone.
- Distinguishing stores by feature allows predictions, for example that acoustically similar words are harder to recall immediately because short-term coding is acoustic.
- The primacy effect is better recall of items at the start of a list, explained by greater rehearsal and transfer to long-term memory.
- The recency effect is better recall of items at the end of a list, explained by those items still being held in short-term memory at recall.
- Middle items are recalled least well, producing a U-shaped serial position curve when recall is plotted against position.
- A delay or interference task before recall reduces the recency effect because the last items are displaced from short-term memory, while primacy is largely unaffected.
- The pattern supports the distinction between short-term and long-term stores in the multi-store model.
- Identifies the serial position curve as better recall of words at the beginning (primacy) and end (recency) of a list than words in the middle.
- Explains primacy as early words being rehearsed and transferred to long-term memory.
- Explains recency as the last words still being held in short-term memory at the point of recall.
- Recognises that Murdock used free recall of word lists of varying length and plotted recall by serial position.
- Links the findings to the multi-store model's distinction between short-term and long-term memory stores.
- Defines reconstructive memory as an active process in which memories are rebuilt at recall rather than simply retrieved unchanged.
- Explains the role of schemas as mental frameworks from past experience that guide interpretation and recall.
- Describes 'effort after meaning' as the mental work of interpreting new material and giving it personal meaning.
- Recognises that reconstruction can lead to distorted, simplified or culturally adjusted memories.
- Links the theory to evidence such as Bartlett's finding that participants changed an unfamiliar story to fit their own schemas.
- Identifies that Bartlett used an unfamiliar Native American folk story, 'The War of the Ghosts', as his stimulus material.
- Recognises that participants reproduced the story repeatedly and their versions changed over time.
- States that participants altered unfamiliar details to fit their own cultural schemas and expectations.
- Explains that Bartlett concluded memory is an active, reconstructive process rather than a passive store.
- Identifies the concept of effort after meaning, where people try to make material meaningful to themselves.
- Defines interference as one memory disrupting another, distinguishing proactive from retroactive interference.
- Explains that context affects accuracy, with better recall when learning and recall contexts match.
- Defines false memories as inaccurate recollections influenced by post-event information or leading questions.
- Applies each factor to a concrete example, such as confusing two similar revision topics or misremembering a witnessed event after discussion.
Examiner Tips
- 💡Underline the key verb in the scenario (taking in, holding, recalling) before choosing an option.
- 💡Remember the sequence input–storage–output and eliminate options that put retrieval before storage.
- 💡Check each option against the definition of one process rather than choosing the longest or most familiar word.
- 💡Ask whether the example is a personal event, a fact, or a skill before selecting an option.
- 💡Watch for distractors that describe a skill but use factual wording; focus on the main content.
- 💡Use a quick three-way check (event, fact, skill) to rule out options systematically.
- 💡Underline the key term in the stem (encoding, storage or a named code) before reading the options.
- 💡Eliminate options that describe a different process from the one asked about, even if the option is a true statement.
- 💡Check each option against the definition you have learned rather than choosing the longest or most technical-sounding answer.
- 💡Match the features in the stem (capacity, duration, encoding) to the store that fits them all.
- 💡Remember the order of transfer: sensory, then short-term, then long-term memory.
- 💡Reject options that describe a store using features belonging to a different store.
- 💡Read the question to identify which feature (coding, capacity or duration) is being asked about before choosing an option.
- 💡Link each feature to a named store, since distractors often pair a correct feature with the wrong store.
- 💡Use the standard approximate figures (7 ± 2 items; 18–30 seconds) as anchors but check the exact wording of the option.
- 💡Identify whether the question concerns the start, end or middle of the list before selecting an answer.
- 💡Remember the U-shape of the serial position curve as a quick check on which positions are recalled best.
- 💡When a delay or interference task is mentioned, predict the effect on recency specifically.
- 💡Read the question stem carefully to see whether it asks about the shape of the curve, the explanation of primacy, or the explanation of recency.
- 💡Eliminate options that describe only one end of the curve when the question asks about the overall pattern.
- 💡Check whether an option confuses short-term and long-term memory before selecting it.
- 💡Look for the key phrase 'active process' in options and reject ones that describe memory as a passive recording.
- 💡When an option mentions schemas, check that it describes them as frameworks from past experience that shape recall.
- 💡Distinguish 'effort after meaning' from general effort or attention by checking whether the option mentions interpreting or making sense of material.
- 💡Link each finding to the concept of schema, since that is the theoretical point of the study.
- 💡Remember the story was unfamiliar to participants, which is why distortions occurred.
- 💡When asked for a conclusion, state that memory is reconstructive and shaped by prior knowledge.
- 💡Use the prefixes pro- (forward, older disrupting newer) and retro- (backward, newer disrupting older) to keep interference types distinct.
- 💡For context questions, state clearly whether contexts match or differ and what effect that has.
- 💡Link false memories to real-world examples such as eyewitness testimony to show application.
Common Mistakes
- Confusing storage with retrieval: the correction is that storage keeps information while retrieval brings it back into awareness.
- Treating encoding as output: the correction is that encoding is the input stage that prepares information for storage.
- Assuming retrieval failure always means the information was never stored: the correction is that poor encoding or a retrieval cue problem can also cause failure.
- Calling a general fact an episodic memory: the correction is that facts and meanings belong to semantic memory.
- Classifying a physical skill as semantic memory: the correction is that how-to skills belong to procedural memory.
- Assuming episodic memory has no link to time or place: the correction is that episodic memories are personal events tied to when and where they happened.
- Confusing storage with retrieval: the correction is that storage keeps information available, while retrieval brings it back into use.
- Assuming all memory is encoded in the same way: the correction is that encoding can be visual, acoustic or semantic depending on the material and store.
- Treating encoding and storage as identical: the correction is that encoding is the initial conversion and storage is the subsequent retention.
- Treating the three stores as one single memory system: the correction is that the model proposes distinct stores with different capacity, duration and encoding.
- Confusing attention with rehearsal: the correction is that attention transfers information from sensory to short-term memory, while rehearsal transfers it from short-term to long-term memory.
- Assuming short-term memory has unlimited capacity: the correction is that short-term memory is limited in both capacity and duration.
- Confusing capacity with duration: capacity is how much is held, duration is how long it lasts. Correction: label each feature explicitly before answering.
- Assuming all stores code semantically. Correction: sensory memory is modality-specific and short-term memory is mainly acoustic, with semantic coding characteristic of long-term memory.
- Treating short-term memory as unlimited. Correction: its capacity is limited to roughly 7 ± 2 items, unlike long-term memory.
- Reversing the effects: primacy concerns the start of the list and recency the end. Correction: link primacy to long-term memory and recency to short-term memory.
- Claiming a delay removes both effects. Correction: a delay mainly removes the recency effect, leaving primacy intact.
- Explaining recency by rehearsal. Correction: recency reflects items still present in short-term memory, not extra rehearsal.
- Thinking the curve shows only the first words are remembered; correction: both the start (primacy) and end (recency) of the list are recalled better than the middle.
- Confusing primacy with recency; correction: primacy refers to the first items (long-term memory) and recency to the last items (short-term memory).
- Assuming participants recalled words in the order presented; correction: Murdock used free recall, so order of output was not fixed.
- Describing memory as a passive store that replays events exactly; correction: reconstructive memory treats recall as an active rebuilding process.
- Confusing 'effort after meaning' with simply trying hard to remember; correction: it refers to the work of interpreting and making sense of material using existing knowledge.
- Thinking schemas are always accurate guides; correction: schemas can cause memories to be distorted to fit expectations.
- Thinking participants remembered the story perfectly: correction is that their reproductions became distorted and shortened over repeated recall.
- Confusing Bartlett's study with a laboratory experiment using nonsense syllables: correction is that he used a meaningful but culturally unfamiliar story.
- Believing the changes were random errors: correction is that changes were systematic, fitting the story to participants' own schemas.
- Reversing proactive and retroactive interference: correction is that proactive means an older memory disrupts a newer one, while retroactive means a newer memory disrupts an older one.
- Treating context as irrelevant to memory: correction is that matching context at learning and recall improves accuracy.
- Assuming false memories require deliberate lying: correction is that false memories are genuine but inaccurate recollections.