Memory
Serial Position Effect
Why we remember the start and the end — but lose the middle
The serial position effect is the U-shaped pattern in free recall: items at the beginning of a list (primacy) and at the end (recency) are remembered better than items in the middle. Hermann Ebbinghaus first noted the curve in his 1885 self-experiments memorizing nonsense syllables. Murdock's 1962 study formalized it. Primacy reflects rehearsal advantage — early items get more rehearsals before list ends. Recency reflects items still in short-term store at recall. The Glanzer and Cunitz (1966) dissociation — delayed recall eliminates recency but preserves primacy — was a key piece of evidence for the multi-store memory model.
- First observedHermann Ebbinghaus (1885)
- Formal studyBennet Murdock (1962)
- Primacy mechanismRehearsal advantage; encoded into long-term memory
- Recency mechanismItems still in short-term/working memory
- DissociationGlanzer and Cunitz (1966) — delay eliminates recency only
- Curve shapeAsymmetric U with stronger recency
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
Why the serial position effect matters
- Eyewitness testimony. First and last events of an incident are best recalled.
- Interviewing. First and last candidates have a memory advantage.
- Lecture design. Mid-lecture content needs reinforcement; openings and closings stick.
- Persuasion. Order of arguments matters — primacy in immediate, recency when delayed.
- Menu design. Top and bottom items get disproportionate attention.
- Cognitive testing. Word-list recall reveals selective memory deficits in dementia.
- Theory of memory. Foundation for the multi-store model and its successors.
Common misconceptions
- The middle is forgotten because it's boring. The pattern emerges with neutral nonsense syllables.
- Primacy and recency share one mechanism. Glanzer-Cunitz showed they dissociate.
- Faster presentation kills primacy. True for primacy (less rehearsal time) but doesn't hurt recency.
- It only applies to word lists. Found in picture, action, abstract, and even autobiographical recall.
- Primacy is just attention. Rehearsal is the dominant mechanism.
- Recency is long-term memory. Recency lives in short-term store; delay erases it.
Frequently asked questions
What does the serial position curve look like?
Plot percent recalled against position in a list of 15-20 items. The first 3-4 items show high recall (primacy), the middle is depressed, and the last 3-4 jump up steeply (recency). Recency is typically larger than primacy. The shape is robust across word lists, picture lists, action sequences, and even abstract concepts. It's one of the most replicated patterns in memory research.
How does primacy work?
When a list is presented, the first item gets rehearsed alone, then with the second, then with the third. By the end, early items have been rehearsed many times; later items few times. More rehearsal means stronger encoding into long-term memory. Atkinson and Shiffrin's (1968) buffer model formalizes this. Slower presentation rates (giving more time per item) increase primacy.
How does recency work?
The last few items are still in short-term/working memory at the moment of recall. They can be unloaded directly without retrieval from long-term store. The capacity is roughly the last 4 items. Anything that disrupts short-term memory between presentation and recall (counting backwards, distractor task) abolishes recency without affecting primacy.
What was the Glanzer and Cunitz dissociation?
Glanzer and Cunitz (1966) ran two conditions: immediate recall (standard U-shaped curve) and delayed recall after 30 seconds of distractor task. Delayed recall preserved primacy (~same as immediate) but eliminated recency. This double dissociation argued for two stores: long-term (primacy, immune to brief distraction) and short-term (recency, washed out by distraction). It was a foundational result for the modal model of memory.
Does primacy need rehearsal?
Mostly yes. Murdock and others showed that suppressing rehearsal (asking subjects to perform a concurrent task) reduces primacy. But not all primacy is rehearsal-based — even when overt rehearsal is prevented, some primacy can survive, suggesting a distinctiveness or attention component. Items presented early occupy a unique temporal context.
Where does this matter outside the lab?
(1) Eyewitness recall — first and last events are remembered best. (2) Job interviews — first and last candidates are remembered better. (3) Lecture design — the middle of a long lecture is most easily lost; chunking and spacing help. (4) Marketing — items at the beginning and end of a list, menu, or shelf get more attention. (5) Persuasion — primacy and recency effects in argument order. The curve isn't just a lab artifact.
Does the effect appear in non-human animals?
Yes. Pigeons, rats, and monkeys show primacy and recency in serial-recognition tasks. This suggests the underlying mechanisms — short-term retention plus rehearsal-like consolidation — are shared across species. Honey bees show some primacy in spatial learning. The curve has thus been used as a comparative tool to study memory architecture.