What Is Working Memory, and Can Training Actually Change It?
What working memory is, how small its capacity really is, and an honest account of why researchers still disagree about whether training transfers to anything.
Topic hub · 22 articles
Working memory is the small amount of information you can hold and manipulate at once, and it is genuinely limited — the classic estimates cluster around four items for most material, not the seven the popular version claims. That limit is why mental arithmetic gets hard as numbers get longer, and why a phone number is easier to hold as three chunks than as ten digits.
What changes with practice is mostly not the limit. It is the strategy: chunking, rehearsal, and encoding material into a form that holds more per slot. A digit span improves largely because the person found a better way to group the digits, which is a real and useful skill even though it is not an expanded container.
These articles cover the span tasks themselves, the encoding strategies that make them easier, the difference between recall and recognition, and what the attention literature says about focus, interference and task switching.
The cornerstone article covers what working memory is, how it is measured, and what training does and does not move.
Chunking and elaborative encoding change performance more than repetition does. The technique articles cover each method concretely.
Most attention failures are interference, not weakness. The attention articles cover what that means for practice design.
What working memory is, how small its capacity really is, and an honest account of why researchers still disagree about whether training transfers to anything.
What attention span actually means, what attention exercises reliably change, what the trials found about transfer, and what to do instead.
Reversing a sequence adds a second job on top of holding it. Why backward span runs shorter than forward span, and what that gap is actually telling you.
How change detection tasks work, what change blindness does and does not prove, and why spotting a removed item depends on what you encoded.
A retention interval is the gap between learning something and being asked for it. Why an empty pause is easy, a filled one is brutal, and what follows.
A digit span is the longest run of numbers you can repeat back. How the stepped procedure produces that figure, and why an app's best span is not a test score.
How grid and position memory tasks work, why location is its own kind of content, and the strategies that help on a grid but not on a list.
Why growing-sequence memory games break down in the middle of the list, what the serial position curve predicts, and how chunking moves the break point.
Concentration is mostly a matter of conditions, not effort. What interruption research shows, and how to set up a session you can actually finish.
Grouping, meaning, phonetic encoding and spatial placement each cost something different. A side-by-side comparison so you can pick the right one.
Forgetting is often competition rather than decay. What proactive and retroactive interference are, how big the effect is, and how switching content helps.
Why your sense of knowing something is a poor predictor of recalling it later, what causes the illusion, and how to grade yourself honestly.
Phonetic, shape, rhyme and peg systems all convert digits into something easier to hold. What each one asks of you before it starts paying anything back.
What an n-back task asks you to do, why dual n-back feels unlike other memory games, and what the training studies actually found about transfer.
What paired-associate learning is, why arbitrary pairs resist repetition, and why judging a pair with both halves in front of you overstates what you know.
Picking an answer from a list is a different job from producing it from nothing. Why the gap exists, and why multiple-choice practice flatters what you know.
Holding a number for thirty seconds is a different problem from memorising it. What the rehearsal loop can carry, and what reliably knocks a number out of it.
One holds information; the other holds it while you do something to it. What separates the two terms, and a quick test for telling which a task needs.
What processing speed means as a psychological construct, how it is measured, and why the response time an app records is a task time rather than a score.
Matching pairs is one memory task among many. The formats we built instead, what each one actually stresses, and the ones that did not survive testing.
Anchoring, chunking and sequence-first methods for holding onto dates, and why memorising isolated years is the least effective approach.
Chunking, the method of loci, elaboration and retrieval practice — which memory techniques have support, and the important limit on what any of them do.