Short-Term Memory and Working Memory Are Not the Same

What the duration figures really mean

"Short-term memory lasts about twenty seconds" is one of those statements that is repeated far more often than it is qualified. Different figures come from different paradigms, and the paradigm is doing most of the work.

Berman, Jonides and Lewis, examining whether time itself is the culprit, are worth quoting on the two halves of the design. They note that rehearsal "happens so habitually that it is often not noticed and is difficult to disengage", which is why experiments block it with a distractor task. And in their comparison of intervals they report that "both conditions with a filled interval led to worse memory performance, but the difficulty of the intervening task had no effect." Their broader conclusion is that the mere passage of time does not cause forgetting in short-term memory, with interference playing the dominant role.

So the honest version is: material survives as long as you can keep refreshing it and as long as nothing competes with it. Take away refreshing and add competition, and it goes quickly. That is a mechanism, not a stopwatch reading, and it is treated more fully in our article on delayed recall and retention intervals.

The capacity is shared, which is why the second job hurts

The reason working memory tasks feel disproportionately harder is that storage and processing draw on the same limited pool. Cowan's review of capacity argues for a central store limited to roughly 3 to 5 meaningful items in young adults, counted in chunks rather than raw items. Every operation you perform on held material competes for that pool.

This is the mechanism behind a familiar experience in mental arithmetic. You know how to multiply 37 by 6. You are not confused about the method. But holding 180 while you work out 42 and then combining them is three live items plus an operation, and if anything interrupts you, the partial result is the thing that vanishes. The method survived; the holding did not. Our guide to mental math methods is largely a collection of ways to reduce that load, and the working memory guide covers what training does and does not change about the pool itself.

How the app's modes fall on either side

IQ Test & Brain Training Games is our own app, built by Reign Creative, and its memory catalogue happens to divide along this line unusually cleanly, which makes it a decent demonstration tool.

On the storage side sit Number Recall and Sequence Memory: the digits appear, they disappear, you reproduce them in order. Nothing is transformed. On the working memory side sit Running Total Memory, where numbers arrive one at a time and you hold the running sum rather than the list; Mental Math Chain, which strings operations together; and the Memory Math mode, which shows numbers and then asks you to calculate with them after they are gone.

Running the two families back to back is the clearest way to feel the difference. Most people find their sequence lengths drop sharply the moment an operation is added, and that drop is the point rather than a failure.

What the app's results are not is a professional assessment. Its memory drills record score, accuracy, streak, response time and personal bests, and it also has a memory test and an IQ-style test that covers working memory; its listing describes those results as estimates based on app activities, not medical, psychological, clinical or professional diagnostic assessments. Interpreting what memory scores mean about a person belongs to qualified professionals using standardised instruments, which is a different activity from practice or entertainment. You can see how the app describes its games and tests on the IQ Test & Brain Training Games app page, explore related writing in the memory and attention hub, and browse the wider education and brain section for the study-method articles.