Serial Position Effect

Category: Memory

Your memory for a list is lopsided: you reliably recall the first few items and the last few items, and the stuff in the middle falls into a hole. Position on the list, not content, decides what survives.

How it works

Free-recall experiments produce a lopsided U-shaped curve: recall probability is high at the start of a list, crashes in the middle, and spikes again at the end. The standard explanation is a two-store account. The final items are still held in short-term (working) memory when recall begins, which produces the recency spike, while the opening items received disproportionate rehearsal and got encoded into long-term memory before the list saturated attention, producing the primacy bump. The cleanest proof that these are two different mechanisms: if you insert a 15-to-30-second distractor task before recall, the recency effect vanishes (short-term store gets flushed) while primacy survives untouched. Middle items lose on both counts, too late for deep rehearsal, too early to still be sitting in the buffer, so they get the worst recall in the list.

Where you'll see it

  • Hiring panels: research on interview order shows candidates seen first and last are rated and recalled far more vividly than the equally qualified people sandwiched in the middle of the day. Your slot on the schedule is doing work your resume should be doing.
  • Restaurant menus and wine lists: designers deliberately place the highest-margin items at the top and bottom of a category, because that is where your eye anchors and your memory holds. The middle is where the stuff they do not care if you order goes.
  • Ballot and survey order: names near the top or bottom of a long list get disproportionate attention and recall, which is exactly why ballot-position laws exist and why survey writers rotate answer options.
  • Cramming for an exam: you review ten topics, walk in, and remember the first two and the last two cold. Topics five through seven, studied just as long, evaporate. You did not study them worse, you studied them in the middle.

Where it comes from

Hermann Ebbinghaus first documented the pattern in his 1885 monograph Uber das Gedachtnis, memorizing lists of nonsense syllables on himself and noticing that middle items took more repetitions to learn than items at either end. The modern quantitative version came from Bennet Murdock in 1962, who had 15 undergraduates free-recall word lists (up to 40 items, one word per second) and plotted the now-classic serial position curve in the Journal of Experimental Psychology. Four years later, Murray Glanzer and Anita Cunitz nailed down the mechanism: in their 1966 paper they showed that a delayed-recall distractor task wiped out recency while leaving primacy intact, giving experimental teeth to the idea of two separate memory stores.

How to counter it

Record before you move on. Do not rely on end-of-session recall to rank a set of items. Score or note each candidate, option, or topic the instant you finish with it, so your judgment is captured before position can distort it.

Interrogate the middle on purpose. When reviewing anything sequential (a shortlist, a menu, a document), consciously go back and re-read the middle. That is precisely where your memory is thinnest and where important things hide unnoticed.

Randomize the order across passes. If you are evaluating items over multiple rounds, shuffle the sequence each time. Any item that consistently wins across different positions is genuinely strong, not just lucky with placement.

Break long lists into chunks. A list of 30 has one primacy peak and one recency peak, leaving 26 items in the dead zone. Five lists of six each create five beginnings and five endings, so far more of the material lands in a well-remembered position.

The tell

You catch yourself defending a choice with "the first one just stood out" or "the last one was clearly the best," and when pressed you cannot actually recall the middle options well enough to say whether they were worse or you simply forgot them.

Related biases

References

  1. Murdock, B. B. (1962). The serial position effect of free recall. Journal of Experimental Psychology, 64(5), 482-488
  2. Glanzer, M., & Cunitz, A. R. (1966). Two storage mechanisms in free recall. Journal of Verbal Learning and Verbal Behavior, 5(4), 351-360
  3. Ebbinghaus, H. (1885). Uber das Gedachtnis: Untersuchungen zur experimentellen Psychologie (Memory: A Contribution to Experimental Psychology). Duncker & Humblot, Leipzig
  4. Kahana, M. J. (1996). Associative retrieval processes in free recall. Memory & Cognition, 24(1), 103-109