Without rehearsal, working memory content decays naturally within roughly ten to twenty seconds

What it is

If a person memorizes a piece of information and is then prevented from silently repeating it or actively maintaining it any other way, correct recall drops sharply within a short window — the classic estimate is that, with no rehearsal and no substantial interfering task, recall falls close to floor within roughly ten to twenty seconds. This finding is specifically about decay under "do nothing, just wait" conditions; it is a different phenomenon from forgetting caused by being interrupted to handle something else in the meantime (which is interference-driven forgetting).

Why it happens

This estimate comes from measuring decay speed after rehearsal is blocked: once the possibility of silently repeating or rehearsing the memorized content is cut off, the ongoing processing needed to maintain it can no longer occur, and its retrievability drops monotonically with elapsed time even when no additional task is inserted. This shows that working memory's "shelf life" is inherently short, and that rehearsal isn't an optional aid — it's one of the necessary conditions for keeping content alive. Without it, the mere passage of time is enough to destroy it.

Studying it

The classic paradigm has participants memorize a short piece of material, then immediately perform a task designed to block silent rehearsal while itself imposing very low cognitive load and no substantial interference (counting backward by threes, for instance), with recall accuracy tested at varying delay intervals to plot a curve of accuracy against elapsed time. A low-load task is chosen specifically to keep "rehearsal blocked" separate from "additional interference introduced" — if the inserted task carries too much load, the measured decay rate gets contaminated by an interference effect and can no longer be attributed to time decay alone.

Where it stops holding

This decay rate was measured under laboratory conditions, with relatively simple material and participants who knew in advance exactly what they needed to remember. The moment any task requiring active processing is inserted, actual forgetting proceeds noticeably faster than this pure-decay estimate, because interference now stacks on top of it. Conversely, if the material itself is easy to organize or connects readily to existing knowledge, decay will be slower than this baseline. This number should be read as a best-case lower bound on decay speed, not a fixed duration that applies to every scenario.

Applying it

• For any flow that asks a user to briefly hold a piece of information and use it directly with nothing else happening in between, design the usable window to fit within roughly ten to twenty seconds; beyond that window, assume the content has already started degrading noticeably rather than trusting that the user still remembers it clearly.
• For verification codes and temporary confirmation codes that must be held briefly, design countdowns and validity windows with the understanding that the content decays naturally even if the user is simply waiting or their attention drifts, with nothing actively interrupting them — don't treat "the user wasn't interrupted" as a guarantee they still remember it.
• How to check: in prototype testing, impose fixed delays (five, fifteen, thirty seconds) before asking users to repeat or re-enter previously shown information, and compare the accuracy drop-off across delays to establish the longest unaided wait the flow can tolerate.

Related

• Same group: A6.02.1 The number of items working memory can hold at once is limited · A6.02.2 Retention is brief and vulnerable to interference · A6.02.3 Requiring users to remember content across pages is itself a design flaw · A6.02.4 The classic capacity estimate is closer to four chunks than the earlier figure of seven · A6.02.6 Rehearsal extends retention but draws on the same limited processing resource · A6.02.7 Capacity varies with item complexity, so simple and complex items can't be compared on one number · A6.02.8 Capacity limits and time-based decay are two independent mechanisms, and easing one doesn't substitute for the other
• Nearby: A6.06 The forgetting curve · A6.19 Prospective memory and forgotten to-dos
• Search terms: Peterson and Peterson task · trace decay · short-term forgetting curve