PSYC10003 · Mind, Brain and Behaviour 1
Memory Systems: Short-Term and Working Memory
Week 4 builds the modal (multi-store) model of memory — encoding, storage, retrieval across sensory, short-term and long-term stores — and characterises short-term memory by its limited capacity (Miller's 7±2, chunking) and brief duration. It then upgrades to Baddeley and Hitch's working-memory model (central executive, phonological loop, visuo-spatial sketchpad, episodic buffer). The examinable contrast, within the 25% Learning & Cognition block, is short-term memory as a passive store versus working memory as active manipulation, plus the serial-position evidence for separate stores.
What this chapter covers
- 01Three memory processes: encoding, storage, retrieval
- 02Atkinson-Shiffrin multi-store model: sensory register → STM → LTM, with rehearsal and retrieval
- 03Sensory memory: iconic (visual) and echoic (auditory); Sperling's partial-report method
- 04STM capacity (Miller's 7±2, chunking) and duration (~15-30 s without rehearsal); Brown-Peterson task
- 05Serial-position curve: primacy (rehearsal → LTM) vs recency (still in STM); the double dissociation
- 06Levels of processing (Craik & Tulving): deeper/semantic encoding → better retention
- 07Baddeley's working memory: central executive, phonological loop, visuo-spatial sketchpad, episodic buffer
- 08STM as passive store vs working memory as active manipulation; neural basis (prefrontal cortex)
Interpreting a serial-position result
- +1Name the effects. Superior recall of the earliest items is the primacy effect; superior recall of the last items is the recency effect. Middle items are worst — too long ago for STM, too little rehearsal for LTM.
- +1Assign the stores. Primacy reflects long-term memory: early items received more rehearsal and were transferred to LTM. Recency reflects short-term memory: the last items are still held in STM at the moment of recall.
- +1Predict the manipulation's effect. A 30-s filled delay (counting backwards prevents rehearsal and lets STM decay) removes the items still in STM, so the recency effect is eliminated.
- +1Explain the selectivity. Primacy is left intact because those early items are already in LTM, not STM. This double dissociation — a filled delay kills recency but not primacy — is the key evidence for two separate stores linked by rehearsal.
Key terms
- Multi-store (modal) model
- Atkinson and Shiffrin's model in which information flows from sensory registers to short-term memory to long-term memory, maintained and transferred by rehearsal; the stores differ in capacity and duration.
- Short-term memory (STM)
- A limited-capacity (about 7±2 items, extendable by chunking), brief-duration (roughly 15-30 s without rehearsal) store holding the conscious present moment.
- Chunking
- Grouping individual items into larger meaningful units to increase effective STM capacity, since capacity is measured in chunks rather than raw items.
- Working memory (Baddeley)
- A multi-component workspace for reasoning and goal-directed cognition — central executive, phonological loop, visuo-spatial sketchpad, episodic buffer — that actively manipulates information rather than merely storing it.
- Primacy and recency effects
- In free recall, better memory for the first items (primacy, via rehearsal into LTM) and the last items (recency, still in STM); their differential vulnerability to manipulations dissociates the two stores.
- Levels of processing
- Craik and Tulving's finding that long-term retention depends on the depth of encoding — semantic (deep) processing yields better memory than phonological or visual (shallow) processing.
Memory Systems: Short-Term and Working Memory FAQ
What is the difference between short-term and working memory?
Short-term memory in the Atkinson-Shiffrin model is a passive, unitary store that holds information briefly. Working memory replaces it with an active, multi-component system that manipulates information for reasoning and problem-solving — the central executive directing the phonological loop and visuo-spatial sketchpad. The exam often hinges on 'passive store vs active manipulation'.
How big is short-term memory really?
The classic estimate is Miller's 7±2 items, but the effective capacity depends on chunking — grouping items into meaningful units — and some estimates are lower (3-4). What counts as an 'item' is itself non-trivial, which is why chunking matters.
What do the phonological loop and visuo-spatial sketchpad do?
The phonological loop holds and rehearses auditory/verbal (speech-based) information; the visuo-spatial sketchpad holds visual and spatial information and supports mental rotation and navigation. They are independent but interacting, which prevents interference between verbal and visual tasks — the basis of dual-task experiments.
Why does deeper processing improve memory?
Because semantic (meaningful) encoding forms richer connections with existing long-term knowledge, giving more retrieval routes later. Craik and Tulving showed recognition was best for words processed for meaning, worse for sound, worst for appearance — shifting the view of STM from a rote buffer to a system that supports meaningful learning.
Is this examined with calculations?
No calculations — it is conceptual. Expect findings-style MCQs: what a serial-position manipulation does, which component of working memory a task loads, or why levels-of-processing produces its ordering. The quantitative memory content is limited to interpreting patterns, not arithmetic.
Exam move
Draw both models from memory — the multi-store flow and the Baddeley working-memory diagram with its four components — and be able to say what each component does. Nail the serial-position logic by pairing each end of the curve with its store and each manipulation with what it attacks (filled delay → recency, concurrent task → primacy). Keep 'passive store vs active manipulation' as your one-line contrast between STM and working memory. For levels of processing, remember the ordering semantic > phonological > visual and the mechanism (richer LTM connections). This corner of the Learning & Cognition block rewards clean conceptual distinctions under MCQ time pressure. Confirm exam details on Canvas.
Working through Memory Systems: Short-Term and Working Memory in PSYC10003? Sia is AskSia’s AI Psychology tutor — ask any PSYC10003 Memory Systems: Short-Term and Working Memory question and get a clear, step-by-step explanation grounded in how PSYC10003 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.
