Serial position effect

The law

People remember the first and the last items in a sequence better than the items in the middle.

Plotted by position, recall forms a U-shaped curve. The two ends of the curve have names:

• Primacy effect: the advantage for the first items.
• Recency effect: the advantage for the last items.

Where it comes from

Hermann Ebbinghaus noted position effects in his self-experiments on memory in the 1880s. Bennet Murdock's 1962 free-recall study, The serial position effect of free recall, measured the curve across lists of different lengths and presentation rates, and the U shape held across all of them.

The usual account treats the two ends as two different mechanisms. Primacy comes from rehearsal: the first items get repeated in the head while there is still little else to rehearse, so they are more likely to move into long-term memory. Recency comes from what is still in short-term memory when recall begins.

Glanzer and Cunitz (1966) showed the split. When people counted for 30 seconds between the list and recall, the recency advantage disappeared while primacy stayed. When items were shown more slowly, giving more time to rehearse, recall of the early and middle items improved while the last items were unaffected.

What it means for an interface

The research used word lists, not screens, so treat it as a tendency to design with, not a rule that predicts a click. It applies wherever people meet items in a sequence and have to hold them in mind:

• Navigation and menus. Put the destinations people need most at the start and the end of a menu or tab bar. Items in the middle of a long menu are the easiest to overlook and to forget.
• Onboarding and explanations. Open with the one thing a person must take away, and close by restating it. The middle of a long tour is where detail gets lost.
• Option lists and plans. When people compare options in a row, the middle ones get the least attention. If the middle option is the one you recommend, mark it explicitly instead of relying on position.
• Research sessions. In interviews and usability tests, participants recall the start and the end of a session best. Ask about the middle tasks while they are fresh, not only at the debrief.

Examples

Tab bars. Mobile apps often put Home first and the account or profile tab last, with less-used destinations in between. The two ends hold the places people return to most.

Onboarding tours. A five-step product tour that opens with the product's main promise and ends by repeating the one action to take next leaves people with that action, even if they forget steps two to four.

Usability debriefs. A participant asked at the end of a session what was hardest will usually name something from the first or the last tasks. Notes taken during the middle tasks catch what the debrief misses.

Where it is misapplied

• Lists people can see. The effect describes recall of items held in memory. When people can see the whole list at once and scan it, position matters less than labels, grouping and visual weight.
• Short lists. Short lists show little middle loss. The problem grows with length, which is an argument for shorter lists before it is an argument for reordering.
• Relying on recency. Recency is fragile: any task between seeing a list and acting on it erases it, as the counting delay did in Glanzer and Cunitz's study.

Checklist

• The most important navigation items sit at the start or the end, not in the middle.
• Explanations open and close with the main point.
• A recommended option in the middle of a row is marked explicitly.
• Long lists are shortened or grouped before they are reordered.
• Research notes capture the middle of a session as it happens.

Related

• Miller's Law: how much of a sequence people can hold in mind at all.
• Peak-End Rule: the end of an experience also weighs most in how it is judged.
• Patterns: Table of Contents.