Cognitive Biases · CB-46
The first and last items in a list or sequence are remembered far better than everything in the middle — a single underlying memory effect with two names for its two ends.
Primacy bias is the tendency to remember the first items in a sequence better than middle items; recency bias is the tendency to remember the last items better than middle items — together forming the well-established 'serial position effect,' in which a plot of recall accuracy against position in a list forms a U-shape, with the middle sagging badly.
The serial position effect was first documented by Hermann Ebbinghaus in memory experiments in the 1880s, and has since been replicated extensively; the distinct contributions of primacy (linked to long-term memory consolidation) and recency (linked to short-term working memory) were clarified through 20th-century cognitive psychology.
The Mechanism
The classic U-shaped recall curve
Ebbinghaus's original memory experiments, and the vast literature since, consistently produce this same U-shape — items presented first benefit from more rehearsal time and consolidation into long-term memory, items presented last are still active in working memory at the moment of recall, and everything in between gets the least of both advantages.
01 · PRIMACY AND RECENCY HAVE DIFFERENT UNDERLYING MEMORY MECHANISMS
This is why they can be affected independently by different manipulations
Primacy effects are linked to long-term memory consolidation (early items get rehearsed more and have more time to transfer into durable memory), while recency effects are linked to short-term working memory (the last items are still active and haven't yet been displaced) — a delay between the list and the recall test reliably weakens recency while leaving primacy relatively intact, demonstrating the two are genuinely separate mechanisms.
02 · IT DIRECTLY SHAPES WHAT PEOPLE REMEMBER FROM MEETINGS, PRESENTATIONS, AND INTERVIEWS
Whatever is said first and last disproportionately sticks
In any sequential presentation of information — a meeting agenda, a sales pitch, a job interview, a list of candidates — items and points made early and late are disproportionately likely to be remembered and weighted in a final judgment, regardless of their actual relative importance compared to points made in the middle.
03 · IT'S DELIBERATELY EXPLOITED IN PERSUASION, NEGOTIATION, AND INTERVIEW ORDERING
Deliberately placing your strongest point first or last is a well-documented strategic move
Skilled communicators and negotiators deliberately place their most important point either first (to benefit from primacy) or last (to benefit from recency), and interview and candidate-ordering research has found that position in a sequence measurably affects evaluators' judgments independent of actual merit — a real source of unfairness in unstructured sequential evaluation processes.
Where It Fails / Inversion
Where it fails / inversion
Not all recall differences by position reflect the bias — sometimes the first or last items genuinely are the most important or memorable on their merits, and content in the middle of a sequence can still be well-recalled if it's made distinctly more vivid, surprising, or emotionally significant than the surrounding material, since raw sequential position is only one of several factors influencing memory.
How To Use It
Worked example · designing a fair candidate interview or evaluation order
Organizations that interview multiple candidates sequentially and then compare them can reduce serial-position distortion by scoring each candidate immediately after their interview (capturing an accurate assessment before it fades or gets overwritten by primacy/recency effects from subsequent candidates), rather than relying on end-of-process recall to compare everyone, which will be disproportionately shaped by whoever was interviewed first and last.
How to use it
When presenting or evaluating a sequence of items, options, or candidates, be aware that whatever comes first and last will be disproportionately remembered and weighted, independent of actual merit — either deliberately place your most important material at those positions, or use a structured, position-independent evaluation method (immediate scoring, written notes) to counteract the effect when fairness matters.
See Also