Nudge Wiki

Concept

Serial Position Effect

Also known as: Primacy and Recency Effect

memory choice-architecture menus

People disproportionately remember and choose from the first and last items in a list — because those are the items best retained in memory (primacy and recency). The practical implication for anyone designing a menu, catalog, or list of options: don't put your least-preferred option first or last, since that position gets disproportionate attention.

Sybil Yang tested this on real restaurant menus with eye-tracking and found a twist: diners scan the whole menu like a book first (matching a straightforward primacy/recency read), but then jump straight to the entrées, pick one, and build the rest of the order around it — meaning entrée placement, not overall page position, deserves the most deliberate design attention (see Yang Menu Eye-Tracking Study). Menus and e-commerce listings ordered purely by ascending price (as Netflix, Spotify, and Uber were noted to do) forgo this lever entirely.

Episode 138 supplies the pricing version of the same lever: putting the most expensive item first on a menu raised average spend 4% in a US bar over eight weeks, with a larger effect in a follow-up on pens (see Lichtenstein Beer-Menu Price-Order Study). First position gets disproportionate attention, so what occupies it should be a deliberate choice rather than a default sort by ascending price — the same complaint made about menus and listings elsewhere on this page.

Order also shapes self-reported wellbeing, not just choice (Episode 196). Students asked "how happy are you with your life?" before "how many dates did you have last month?" reported roughly five times higher happiness than students asked the same two questions in reverse order — the dating question primed a narrow, often-unflattering frame that carried over into the broader happiness judgment (see Dating-Happiness Question-Order Study).

Applied to marketing and hospitality (Episode 82). Katelyn Bourgoin cites a courtroom-verdict finding that information presented later in a trial disproportionately sways verdicts compared to earlier testimony, and connects it to Dishoom's dice-roll free-meal finale (see Dishoom Matka Dice-Roll Case) as a deliberate final-impression play — the same mechanism the Octopus Energy wheel-of-fortune case (see Octopus Energy Wheel-of-Fortune Meter-Reading Case) was independently framed around before its later documented citation (Episode 296) treated it as a variable-reward example instead; both framings describe the same real Octopus feature.

A classic demonstration with confabulated reasoning attached (Episode 251). A 1977 study by Timothy Wilson and Richard Nisbett had participants evaluate four physically identical pairs of tights shown one after another, then explain their choice. Preference rose steadily with position — 12% picked the first pair, 17% the second, 30% the third, 40% the fourth — a pure recency-driven serial-position effect, since later items could be compared against those already seen. No participant cited position as a factor; all confidently invented specific (false) sensory reasons for their preference (see Nisbett Wilson Pantyhose Serial-Position Study (1977)). Cited by Paul Zak as evidence people can't reliably introspect on the real causes of their own choices — see Neurologic Immersion.

Discussed in

  • Episode 6 — 6- The psychology behind effective pricing
  • Episode 82 — 3 Nudges I Wish I Learned at Business School (Dishoom dice-roll finale; Octopus Energy wheel-of-fortune cross-link)
  • Episode 138 — 138-listen-to-exactly-17-minutes-and-42-seconds-of-this-episode
  • Episode 196 — 196-7-cognitive-biases-that-might-ruin-your-life
  • Episode 251 — 251-how-your-brain-can-reveal-what-you-buy (Nisbett/Wilson pantyhose study; confabulated reasoning)
  • Episode 252 — 252-new-coke-and-the-marketing-blunder-of-the-century (pantyhose study re-cited by Philip Graves)

See Primacy Effect for the fuller treatment of first-position advantage across judgment, memory, and real-world elections, and its recency-effect counterpart.

Related