Information encoded by meaning and linked to what you already know is much more likely to be stored for a long time.
Memory · processes of memory
Storing and finding memories
Long-term memories are stored in an organised way, linked by meaning, a bit like a filing system.
Retrieval works best with a cue: a place, a smell, a picture or the first letter of a word.
A memory can be stored but not found. "On the tip of my tongue" is a retrieval problem, not a lost memory.
Quick check
?Ayo learns a door code by saying it over and over under his breath. How is he encoding it?
Quick check
?Zara reads a page of her textbook while thinking about the weekend. The next day she can't recall a single fact from it. Which process most likely failed?
Your answer
?You know your old teacher's name and you recognise it when a friend says it, but you can't bring it to mind yourself. Which process has failed?
Memory · processes of memory
Three types of long-term memory
Episodic memory: memory of personal events, with a time and place, e.g. your last birthday.
Semantic memory: memory of facts and knowledge, e.g. Paris is the capital of France.
Procedural memory: memory of how to do things, e.g. riding a bike. It is used without conscious effort and is hard to put into words.
Sort it
?Which type of long-term memory is each person using?
Tap a statement, then tap the group it belongs in. Then press Check.
Mission check
?A number's gone in seconds; a joke lasts years. What's different?
You held the number only as a sound, for a few seconds. The joke you understood: you encoded it by its meaning and linked it to things you already knew, so it was stored in long-term memory and could be retrieved years later.
Mission 2
You see everything for a split second. What decides what you keep?
Make a guess: You see everything for a split second. What decides what you keep?
Memory · the multi-store model
Three stores in a line
Atkinson and Shiffrin's (1968) multi-store model says memory is made of three separate stores.
Information flows from the sensory register to short-term memory (STM) to long-term memory (LTM).
Attention moves information from the sensory register into STM. Anything not attended to is lost.
Rehearsal (repeating it) keeps information in STM; enough rehearsal transfers it to LTM.
To use a memory, it is retrieved from LTM back into STM.
Memory · the multi-store model
Encoding, capacity and duration
Each store differs in how it encodes information, how much it holds and how long it lasts.
Store
Encoding (coding)
Capacity
Duration
Sensory register
Same form as the sense (images, sounds)
Very large
Under a second for sights, about 2 seconds for sounds
STM
Mainly acoustic (sound)
About 7 ± 2 items
About 18–30 seconds without rehearsal
LTM
Mainly semantic (meaning)
Unlimited
Potentially a lifetime
Quick check
?Mia hears a 12-digit order number once. When she tries to repeat it straight away, she gets only the first seven digits. Which feature of STM explains this?
Memory · the multi-store model
Primacy and recency
In free recall of a list, how likely you are to recall a word depends on its serial position (where it came in the list).
Primacy effect: the first words are recalled well. They were rehearsed most, so they reached LTM.
Recency effect: the last words are recalled well. They are still in STM at recall.
Middle words are recalled worst: pushed out of STM before enough rehearsal.
Key studyMemory · the multi-store model
Murdock (1962)
Aim
To find out whether a word's position in a list affects how likely it is to be recalled.
Method
Lab experiment. Participants heard lists of common words (10 to 40 words, one word every 1 or 2 seconds), then recalled as many as they could in any order.
Results
Words at the start and end of each list were recalled best and words in the middle worst, giving a U-shaped serial position curve.
Conclusion
The primacy effect (from LTM) and the recency effect (from STM) support separate short-term and long-term stores.
Strength
A highly controlled, standardised procedure, so it is easy to replicate.
Weakness
Learning word lists is artificial, so it may not reflect everyday memory (low ecological validity).
Quick check
?Students hear a list of 20 words. Group 1 recalls straight away. Group 2 counts backwards for 30 seconds first. Group 2 recalls the first words as well as Group 1 but loses the last words. Why?
Sort it
?Is each point a strength or a weakness of the multi-store model?
Tap a statement, then tap the group it belongs in. Then press Check.
Mission check
?You see everything for a split second. What decides what you keep?
Attention decides. Your sensory register holds almost everything for a split second, but only what you pay attention to passes into short-term memory; the rest is lost. Rehearsal then decides what makes it into long-term memory.
Mission 3
Retell a strange story over months and it changes. Who edits it?
Make a guess: Retell a strange story over months and it changes. Who edits it?
Memory · reconstructive memory
Memory is rebuilt, not replayed
Memory is not a recording. Each time you recall something, you reconstruct it from stored fragments plus what you know.
This is Bartlett's (1932) Theory of Reconstructive Memory: memory is an active process.
Gaps are filled with what seems likely, so memories change without you noticing.
Effort after meaning: we make sense of new information by fitting it to what we know, changing it as we do.
Memory · reconstructive memory
Schemas
A schema is a mental framework of knowledge and expectations about something, e.g. what happens at a birthday party.
How schemas form: through experience and culture (repeated events, what we see and are told). New experiences update them.
How schemas shape memory: we recall what fits, leave out what doesn't, turn the unfamiliar into the familiar and add details we expect.
Key studyMemory · reconstructive memory
Bartlett (1932)
Aim
To see whether people's schemas, built from their culture, change how they recall an unfamiliar story.
Method
20 English university students read a Native American folk tale, "The War of the Ghosts", twice, then recalled it after 15 minutes and repeatedly over months and years (repeated reproduction).
Results
Recalls became shorter; strange and supernatural details were left out or explained away; details changed to fit English culture, e.g. "canoes" became "boats" and "hunting seals" became "fishing".
Conclusion
Memory is reconstructive: people use schemas to fill gaps and make an unfamiliar story make sense to them.
Strength
A meaningful story is closer to everyday memory than lists of words.
Weakness
Participants weren't told to be accurate, so some changes may be guesses; small, one-culture sample.
Quick check
?Volunteers wait two minutes in a dentist's waiting room, then are asked to list what was in it. Several list a fish tank and a pile of magazines. There was a fish tank, but no magazines. What best explains the magazines?
Sort it
?Is each point a strength or a weakness of the Theory of Reconstructive Memory?
Tap a statement, then tap the group it belongs in. Then press Check.
Mission check
?Retell a strange story over months and it changes. Who edits it?
Your own mind edits it, without you noticing. Each time you recall the story you rebuild it using your schemas, so strange details are dropped or changed into familiar ones, and the story becomes shorter and more like what you already know.
Mission 4
Same crash, but asked "hit" or "smashed". Why do memories differ?
Make a guess: Same crash, but asked "hit" or "smashed". Why do memories differ?
Memory · accuracy of memory
Leading questions
A leading question suggests the answer, e.g. "Did you see the knife?" rather than "Did you see a knife?"
Information met after an event (questions, other people's stories, news) can blend into your memory.
The memory is distorted, and the witness doesn't realise.
In experiments, asking how fast cars "smashed" rather than "hit" led to higher speed estimates, and more people later "remembered" broken glass that wasn't there.
Memory · accuracy of memory
Schemas and witnesses
Witnesses don't take in every detail, so they fill gaps from their schemas: what usually happens, or what a "typical" robber looks like.
This can add details that never happened, or change ones that did.
Eyewitness accounts can decide court cases, so these errors matter.
A witness can be confident and still wrong.
Memory · accuracy of memory
False memories
A false memory is a memory of something that didn't happen, or happened differently. It feels just as real as a true memory.
False memories can be created by suggestion: leading questions, other people's accounts, imagining an event, or images and adverts.
Each time a memory is rebuilt, new information can get into it.
Quick check
?A police officer wants an accurate account from a witness to a car accident. Which question is most likely to distort the witness's memory?
Memory · accuracy of memory
Interference
Interference: other information gets in the way of recall, especially when it is similar.
Proactive interference: old learning disrupts new, e.g. you keep giving your old phone number when asked for the new one.
Retroactive interference: new learning disrupts old, e.g. after learning Spanish words, the French ones you learned before are harder to recall.
Memory · accuracy of memory
Context
Recall is better in the same context as learning: the same place, situation or mood.
The context acts as a cue that triggers the memory.
So you forget why you came upstairs, then remember when you go back downstairs.
Revising in a very different setting from the exam can make recall harder.
Sort it
?What best explains each memory error?
Tap a statement, then tap the group it belongs in. Then press Check.
Sort it
?Is each point a strength or a weakness of research into leading questions and false memories?
Tap a statement, then tap the group it belongs in. Then press Check.
Mission check
?Same crash, but asked "hit" or "smashed". Why do memories differ?
The question changed the memory. A leading word like "smashed" is information after the event, and it gets blended into the memory when it is rebuilt, so the witness may now recall a faster, more violent crash, and even details that weren't there.
Recap
Big ideas
Memory has three processes: encoding (input), storage and retrieval (output). There are three types of long-term memory: episodic, semantic and procedural.
Information can be encoded acoustically, visually or semantically; encoding by meaning makes long-term storage much more likely, and cues help retrieval.
The multi-store model has three stores, sensory register, STM and LTM, which differ in encoding, capacity and duration.
Attention moves information into STM; rehearsal keeps it there and transfers it to LTM.
Primacy and recency effects (Murdock's serial position curve) support separate STM and LTM.
Memory is an active process: we reconstruct memories from fragments and our schemas, so recall can change without us noticing.
Schemas form through experience and culture and shape what we recall: we leave out, change and add details to fit what we expect.
Bartlett called this "effort after meaning", and his War of the Ghosts study showed recalls becoming shorter and more familiar.
Memory can be distorted after the event: leading questions and other information blend into the memory, and schemas fill gaps, sometimes creating false memories that feel real.
Accuracy is also affected by interference (proactive and retroactive) and context (recall is better in the same context as learning).
🏆
Mini-lesson complete!
⭐⭐⭐
You've worked through Memory for AQA GCSE Psychology (8182). 🎉
Mission 1A number's gone in seconds; a joke lasts years. What's different?
Mission 2You see everything for a split second. What decides what you keep?
Mission 3Retell a strange story over months and it changes. Who edits it?
Mission 4Same crash, but asked "hit" or "smashed". Why do memories differ?