IIF Certified Assessment
Start IQ Test
IIF Certified Assessment
The correlation between them is one of the sturdier findings in cognitive psychology. How strong it is, and what it implies about training either one, is where the agreement stops.
Working memory is one of the few technical terms in cognitive psychology that has escaped into ordinary speech, usually in a slightly wrong form. It does not mean short-term storage, and it is not a measure of how many things you can remember. It means holding information available while operating on it — keeping the first half of a sentence live while parsing the second, tracking two constraints while checking whether a third is consistent with them.
The standard laboratory tools are complex span tasks. You are given something to remember, then interrupted with an unrelated judgement, then given the next thing to remember, and so on, and finally asked to recall the sequence. The interruption is the point: simple storage would be easy, and the task is built so that storage and processing compete.
Clinical batteries measure something adjacent. The Wechsler scales, for instance, carry a working memory index built from subtests such as repeating digit sequences forwards and backwards. It is not identical to a laboratory complex span task, which is one reason effect sizes differ across studies that all say “working memory”.
Reasoning, in this literature, usually means fluid reasoning: solving novel problems that do not depend on stored knowledge — matrix items, series completion, logical inference from unfamiliar premises. Across a large number of studies, working memory measures and fluid reasoning measures correlate substantially and positively. That finding is robust and has been replicated many times.
What the correlation means is where the field divides. One camp reads it as evidence that the capacity to maintain and control information is close to the core of what reasoning ability is. Another argues the overlap is real but has been inflated by shared task demands — both kinds of test require you to sit still, follow instructions, resist distraction and work under time pressure — and that stripping those shared demands out leaves two related but genuinely distinct abilities.
Two things can be strongly related and still not be the same thing. That gap is where the argument lives.
This is not an academic dispute, because a whole product category rests on the answer. If working memory capacity is close to the core of reasoning, and if it can be trained, then training it should raise reasoning ability. That syllogism is the entire premise of the brain-training industry.
The evidence has been unkind to it, and specifically to the second and third steps. Training on a working memory task reliably improves performance on that task, sometimes dramatically. Improvement on closely similar tasks — what researchers call near transfer — appears sometimes. Improvement on genuinely different measures, including reasoning tests, is what is called far transfer, and it is the claim that has repeatedly failed to hold up in larger, better-controlled studies and in meta-analyses that account for the quality of the control condition.
A full assessment reports more than one number for exactly this reason: reasoning, working memory and processing speed are related but separable, and a single total can hide a profile in which they diverge sharply. A result worth acting on names its scale, gives the percentile against a defined reference sample, shows the confidence range, and breaks the total down into the domains behind it.
Find your IQ score now! →The useful takeaway is not a number. It is that “related” and “the same” are different claims, that most popular writing collapses them, and that a product built on the collapsed version is selling a step the research has not delivered.
The ability to hold information in mind while doing something with it — not simply storing it. Keeping the first half of a sentence live while you parse the second half is working memory; reciting a phone number you just heard is closer to simple short-term storage.
No. Measures of the two correlate substantially, which is well replicated, but they are not interchangeable and the size of the relationship varies with how each is measured. Treating them as the same construct goes further than the evidence supports.
They reliably improve performance on the tasks they train. Whether that carries over to general reasoning — far transfer — is not established, and studies using an active control group, where the comparison group does something equally engaging, consistently report smaller effects than studies without one.
Corrections: spotted an error? Email corrections@iqmetrics.org and we will update this story and note the change here.
Three rows, three columns, two things changing at once. The worked solution is below — and so is an honest account of what solving it does and does not tell you about yourself.
A bad night does not lower intelligence. It reliably lowers the things a timed test happens to measure most closely — sustained attention, working memory and speed — while leaving vocabulary and general knowledge comparatively intact.
Average scores climbed for decades, then flattened and in places slipped. None of it shows on a score report, because every test is reset so the average is 100 again — which is why a 1990 score is not a 2026 score.
Our IIF-certified assessment reports your score with its scale, percentile and confidence range — and a breakdown of the cognitive domains behind it.
Start IQ Test →
