This question splits into two that get tangled together, and almost every argument about brain training comes from treating them as one.
Can you get better at IQ tests? Yes, clearly and fairly easily. Can you raise your general reasoning ability, the thing the test is trying to estimate? That is a much harder question, and the honest answer is that fifteen years of serious effort has produced very little evidence that you can, outside of one notable exception.
The study that started it
In 2008 a team led by Susanne Jaeggi published a result that electrified the field. Participants trained on a demanding working memory task called dual n-back, and afterwards scored higher on a test of fluid intelligence. The improvement was dose dependent: more training days, bigger gain (Jaeggi et al., 2008).
If that held, it would have been one of the most useful findings in psychology. Fluid intelligence was supposed to be the part that training does not touch. A free task you could do on a laptop appeared to be moving it.
An entire industry grew out of that paper and the ones around it. It is worth being clear that the researchers themselves were more careful than the marketing that followed.
What happened next
Other labs tried to reproduce it, and the result did not hold up well.
The most pointed test came from Redick and colleagues in 2013. They ran a randomised, placebo-controlled study, which matters enormously here: if your control group does nothing while your training group does something engaging for weeks, you have measured expectation as much as cognition. With a proper active control, they found no evidence of intelligence improvement after working memory training (Redick et al., 2013).
In the same period, Melby-Lervag and Hulme pooled the available studies in a meta-analysis. Their finding set the shape of the current consensus. Training produces reliable short-term gains on the trained task and on tasks closely resembling it. It does not produce convincing evidence of transfer to general ability (Melby-Lervag and Hulme, 2013).
Researchers call this the near transfer and far transfer distinction, and it is the most useful idea in this whole area. Practise something and you get better at it. Get better at something slightly different as a side effect, sometimes. Get better at reasoning in general, almost never.
The exception, and it is a large one
One intervention does appear to work, and it is not an app.
Ritchie and Tucker-Drob pooled studies using three different research designs, including natural experiments where schooling length changed for reasons unrelated to the pupils, such as a change in the law. Every design pointed the same way: an additional year of education was associated with a gain of roughly one to five IQ points, and the gain persisted rather than fading (Ritchie and Tucker-Drob, 2018).
That is an odd result to sit alongside the brain training failures, and the likely reason is duration and breadth. A year of school is thousands of hours across every kind of material, with feedback, in a social setting, at an age when a great deal is still forming. Twenty minutes a day of one puzzle for a month is not a small version of that. It is a different thing. What this means for children specifically is worth its own treatment.
Practice effects are real, and they are not cheating
Separately from all of this, you will score higher the second time you take a test. This is well established and it is one of the reasons a single score should never be treated as final.
The gain comes from knowing the format, recognising item types, not wasting attention on the instructions, and being less nervous. On a matrix test the effect is noticeable, and it is largest for people who have never seen that kind of puzzle before.
That is not fraud, but it is worth naming for what it is: the test is now measuring something slightly different, because part of what it measured the first time was unfamiliarity.
| What people try | Does it raise your score on a test? | Does it raise general ability? |
|---|---|---|
| Practising the exact test you will sit | Yes, several points | No |
| Working memory or n-back training | Little or none | No convincing evidence |
| Commercial brain training apps | On the app’s own games | No |
| Staying in education longer | Yes | Yes, by the best available evidence |
| Sleep, food, not being ill on the day | Yes, recovers points you would have lost | No, it removes a handicap |
| Listening to Mozart | No | No |
I tried this myself
I spent six weeks on dual n-back a few years ago, partly out of professional curiosity and partly because I am competitive about things I should not be competitive about. Twenty minutes most evenings.
I got dramatically better at dual n-back. I went from losing the thread at level two to holding level four fairly reliably, and I was quietly delighted with myself. Then I sat a matrix reasoning test and scored almost exactly what I had scored before starting.
What I did notice, and this is the part I think is actually useful, is that after several weeks of matrix-style puzzles the questions began to feel familiar in a specific way. I started recognising the families: this one is rotation, this one has two rules running at once, this one is counting. That recognition genuinely made me faster and more accurate on tests. It did not make me a better thinker. It made me an experienced test taker, which is a real and much narrower thing.
What to do with this
If you want a higher number on a test, practise that format. It works, within a few points, and there is nothing dishonest about it as long as you know that is what you are buying.
If you want to be a better reasoner, the evidence points at education, at sustained engagement with difficult material over years, and at not being tired, hungry or ill when you are trying to think. That is a duller answer than an app, and it is the one the research supports.
And if you simply want a current reading, take one properly rather than repeatedly. The free IQ test for adults takes about fifteen minutes, and the children’s version is age-adjusted. Before reading much into the result, it helps to know how a raw score becomes an IQ figure and what the percentile actually represents.
One last thing that puts this in perspective. Average scores across whole countries rose for most of the twentieth century, by far more than any training study has ever produced, and then in several countries they started falling again. That story says more about what moves intelligence than any brain training trial does. And if you are wondering how ability changes over a lifetime without any intervention at all, the pattern across adulthood is not the simple decline most people expect.
References
- Jaeggi, S. M., Buschkuehl, M., Jonides, J., & Perrig, W. J. (2008). Improving fluid intelligence with training on working memory. Proceedings of the National Academy of Sciences, 105(19), 6829-6833. doi:10.1073/pnas.0801268105
- Redick, T. S., Shipstead, Z., Harrison, T. L., Hicks, K. L., Fried, D. E., Hambrick, D. Z., Kane, M. J., & Engle, R. W. (2013). No evidence of intelligence improvement after working memory training: A randomized, placebo-controlled study. Journal of Experimental Psychology: General, 142(2), 359-379. doi:10.1037/a0029082
- Melby-Lervag, M., & Hulme, C. (2013). Is working memory training effective? A meta-analytic review. Developmental Psychology, 49(2), 270-291. doi:10.1037/a0028228
- Ritchie, S. J., & Tucker-Drob, E. M. (2018). How much does education improve intelligence? A meta-analysis. Psychological Science, 29(8), 1358-1369. doi:10.1177/0956797618774253

