You are viewing a free preview of this lesson.
Subscribe to unlock all 10 lessons in this course and every other course on LearningBro.
The models studied so far — the multi-store model, the working memory model, Tulving's types of long-term memory — describe memory as though it worked identically in everyone. In reality, memory ability varies enormously between people, and cognitive psychology's individual-differences perspective sets out to describe and explain that variation. Some of it is developmental (memory changes systematically from infancy to old age); some reflects atypical cognitive profiles associated with conditions such as dyslexia and autism; and some reflects expertise — the striking finding that experts remember material in their own field far better than novices, not because they have a "better memory" in general, but because they organise information differently. This lesson examines these sources of variation and their consequences, including for the accuracy of eyewitness memory, always insisting that claims be grounded in real evidence rather than stereotype.
Key Definition: An individual-differences approach to memory studies how and why memory performance varies systematically between people — as a function of developmental stage, atypical cognitive profile, ageing, or domain expertise — rather than treating memory as a single, uniform faculty.
This lesson addresses the Edexcel 9PS0 — Paper 1, Topic 2: Cognitive Psychology content on individual differences in memory: how memory ability varies with developmental level, with atypical profiles such as dyslexia and autism, and with ageing, and how expertise and chunking produce large, domain-specific advantages. It applies the topic's core constructs — the phonological loop and central executive of the working memory model, chunking (Miller, 1956), and Tulving's types of LTM — to explain why people differ. In assessment-objective terms, you should be able to describe the principal sources of individual variation in memory (AO1), apply them to explaining differences between named individuals or groups (AO2), and evaluate the evidence — its methodology, its tendency to generalise, and the danger of deficit-focused stereotyping (AO3).
Connects to…
The single most robust source of individual difference is developmental level: memory performance improves markedly across childhood and adolescence. Two well-evidenced trends matter for the exam.
First, basic capacity increases. Digit span — the number of digits a person can repeat back in order — rises steadily through childhood: a typical four-year-old manages around four digits, whereas by adolescence performance approaches the adult span of around seven (as first quantified by Jacobs, 1887, whose serial-digit technique remains the standard measure). This growth is closely tied to the maturation of the phonological loop of the working memory model. As the speed of sub-vocal rehearsal (articulation rate) increases with age, more items can be refreshed before the phonological store's trace decays, so the effective span expands. This is why span and speech rate are correlated across development — a point demonstrated directly in the contemporary study examined later in this topic.
Second, children increasingly use memory strategies. Younger children tend not to spontaneously rehearse or organise material; as they mature they begin to rehearse, then to group items by meaning, and finally to use elaborative strategies — a progression covered fully in the next lesson. The upshot is that developmental differences reflect both a growing hardware capacity (a faster loop, a more capable central executive) and the acquisition of better software (strategies), and the strongest answers distinguish the two.
| Developmental change | Underlying mechanism | Consequence for memory |
|---|---|---|
| Faster articulation rate | Maturation of the phonological loop | Longer digit span; more verbal material held |
| Better attentional control | Maturation of the central executive / prefrontal cortex | Less distractibility; better manipulation of information |
| Strategy acquisition | Learning to rehearse, organise and elaborate | More is transferred into LTM; recall improves |
| Growing knowledge base | Expanding semantic memory | More effective chunking; new material integrated more easily |
Dyslexia is a specific learning difficulty affecting reading and spelling that is not explained by low general intelligence or poor teaching. The dominant cognitive account — the phonological deficit hypothesis — locates the core difficulty in the processing of speech sounds (phonology), and this connects dyslexia directly to the memory topic through the phonological loop.
Key Definition: The phonological deficit hypothesis proposes that the core difficulty in dyslexia is a weakness in representing, storing and retrieving the sound structure of language (phonology), which impairs the phonological loop and, downstream, the learning of letter–sound correspondences needed for reading.
People with dyslexia frequently show a reduced verbal short-term memory span and difficulties on tasks that load the phonological loop — for example, remembering sequences of digits or repeating unfamiliar non-words (a non-word repetition task is one of the most reliable markers). The interpretation is that a less efficient phonological store or articulatory process holds fewer speech-based items, which makes it harder to hold a word's sounds in mind while decoding it letter by letter. Crucially, the difficulty is domain-specific: visuo-spatial short-term memory (the sketchpad) is typically intact, and many people with dyslexia have strong or superior visual and spatial reasoning. This dissociation — impaired verbal, spared visual short-term memory — is itself evidence for the separable components of the working memory model, and it is a reason the WMM is preferred to the multi-store model's single, undifferentiated STM.
It is important to state the evidence carefully. The association between dyslexia and phonological-loop weakness is well replicated, but dyslexia is heterogeneous: not everyone with a diagnosis shows the same profile, and competing accounts (a rapid-temporal-processing deficit, a cerebellar/automaticity deficit, a visual-attention-span deficit) each capture part of the picture. The defensible teaching point is that dyslexia is associated with a phonological-loop difficulty for many individuals — not that every person with dyslexia has an identical memory profile.
Autism spectrum condition (ASC) is associated with a distinctive — and notably uneven — memory profile rather than a global memory impairment. The research supports several reasonably robust patterns.
The key exam point is that autism produces a peaks-and-troughs profile — enhanced rote/detail memory, reduced gist-based distortion, but relative episodic and strategic difficulty — which maps neatly onto the topic's constructs (Tulving's episodic/semantic distinction; Bartlett's schema-driven reconstruction). As with dyslexia, this must be stated with care: autism is highly heterogeneous, the findings describe group tendencies with wide individual variation, and older studies used small, unrepresentative samples.
| Profile | Verbal (phonological) STM | Visuo-spatial STM | Gist / reconstructive tendency | Episodic memory |
|---|---|---|---|---|
| Typical developmental | Age-typical | Age-typical | Strong (gist-driven; prone to schema distortion) | Age-typical autonoetic recall |
| Dyslexia (common profile) | Reduced | Typically spared / strong | Typical | Typical |
| Autism (group tendency) | Often age-typical | Often strong for detail | Reduced (detail-focused; less gist distortion) | Relative difficulty; less autonoetic |
Memory change in older age is one of the best-documented individual differences — and one of the most misunderstood, because it is highly selective. Ageing does not erode all memory equally.
A further robust finding, from Bahrick et al. (1975), qualifies the decline picture for very well-learned material: recognition of former classmates' faces from yearbook photographs remained around 90% even decades after leaving school, showing that highly over-learned long-term memories are remarkably durable across the lifespan. The distinction that matters is therefore not "young good / old bad" but which system and which task: fluid, effortful, self-initiated retrieval declines, while crystallised knowledge and well-consolidated memories endure.
flowchart TD
AGE["Healthy ageing"]
AGE --> DEC["Tends to DECLINE"]
AGE --> PRES["Tends to be PRESERVED"]
DEC --> EP["Episodic memory<br/>(recent events, free recall)"]
DEC --> WM["Working memory<br/>(central executive, dual-tasking)"]
DEC --> PRO["Prospective memory<br/>(remembering to act)"]
PRES --> SEM["Semantic memory<br/>(facts, vocabulary — may grow)"]
PRES --> PROC["Procedural memory<br/>(well-learned skills)"]
PRES --> OVER["Over-learned LTM<br/>(Bahrick et al., 1975)"]
Perhaps the most theoretically important individual difference is expertise. Experts recall vastly more material in their domain than novices — but not because their raw memory stores are larger. The mechanism is chunking (Miller, 1956): experts have, through years of practice, built up a huge store of meaningful patterns in long-term memory, so they can group incoming information into a few large, meaningful chunks rather than many small ones.
The classic demonstration is Chase and Simon's (1973) study of chess. Presented briefly with a chessboard from a real game, chess masters reproduced it far more accurately than novices — apparently confirming a "better memory". But the crucial control was a board with the same number of pieces placed at random: here the masters' advantage vanished, and they performed no better than novices. The masters were not remembering individual pieces; they were recognising familiar configurations (an opening, a known attacking formation) and recoding many pieces into a few meaningful chunks. Randomising the pieces destroyed the meaningful patterns, removing the basis for chunking. This is decisive evidence that the expert advantage is knowledge-based and domain-specific, not a general superiority of the memory system — an expert chess player has an ordinary digit span.
Key Definition: Expertise-based chunking is the use of extensive domain knowledge held in long-term memory to group incoming information into a small number of large, meaningful units, dramatically increasing the amount of domain-relevant material that can be held and recalled — with no change to general memory capacity.
The same effect has been shown across domains: expert readers of a language chunk letters into words and phrases; experienced medical staff recall a patient's presentation as a small number of clinically meaningful patterns; skilled musicians remember long passages as structured phrases. A related classic case is Ericsson and colleagues' demonstration that an ordinary participant, "SF", trained his digit span from around 7 up to about 79 digits over many hours of practice — achieved entirely by inventing a chunking scheme (encoding digit groups as running times, since he was a keen runner) rather than by expanding STM itself. When switched to letters, his span dropped back to normal, proving the gain was a domain-specific strategy, not a bigger store. Expertise, then, does not raise the ceiling of short-term memory; it changes what counts as a single "item".
These sources of variation converge on a practically important consequence: eyewitness memory accuracy is not uniform — it depends on who the witness is. (Note that the full study of eyewitness testimony and the cognitive interview belongs to criminological psychology, not here; this section addresses only the individual-differences angle.)
Subscribe to continue reading
Get full access to this lesson and all 10 lessons in this course.