Does Brain Training Work? What the Research Shows

    Memory Training

    Does Brain Training Work? What the Research Shows

    Brain-training games make you better at the games. Whether that spreads to the thinking you do every day is the real question. What Jaeggi 2008, Melby-Lervåg and Hulme 2013 and Simons 2016 found, what they do not show, and how to train the memory you actually use.

    By Published Updated

    If you have ever played a brain-training game for a few weeks, you will know the feeling. The scores climb. New levels unlock. The little graph goes up and to the right. Then you walk into the kitchen and cannot remember what you came in for.

    So, does brain training work? The research gives a clear answer with a twist in it. Yes, you get better at the games you practise. That part is solid. What the evidence does not show is that the improvement spreads much further: to your reasoning, your intelligence, or the thinking you do in everyday life. A major 2016 review of the studies that brain-training companies themselves cite found little evidence of that.

    The core lesson is simple: you get better at what you practise, and the gain stays close to what you practised.

    Near and far: the question that matters

    Two terms make sense of the whole debate.

    Near transfer is getting better at the task you trained, or at tasks that look very like it. Practise a matching game and you get better at matching games.

    Far transfer is the big promise: practise a game and get better at something quite different, such as reasoning through a problem you have never met, following a long conversation without losing the thread, or remembering the names at a dinner party.

    Brain training sells far transfer. A few minutes a day on a game, and a sharper mind everywhere else. That is the claim worth testing.

    The study that started the boom

    The research: Jaeggi et al, 2008

    • In short: People who trained on a demanding working memory game (working memory is the small mental workspace you hold things in while you think), the dual n-back, then scored higher on tests of fluid intelligence, and the more they trained, the bigger the gain.
    • What they did: Participants practised the dual n-back for different amounts of time. In the dual n-back you take in a stream of sounds and positions at once, and for each one you say whether it matches the one a set number of steps back. Fluid intelligence, the ability to reason through new problems, was tested before and after training.
    • Who was involved: Adults, some training on the dual n-back for different lengths of time, compared with untrained controls.
    • What they found: Fluid intelligence scores rose after training, and the gain increased with the amount of training.
    • What they concluded: The authors concluded that working memory training could improve fluid intelligence.

    What kind of evidence is this? A single training study, published in PNAS, a leading journal, and widely cited. It is the paper that set off the brain-training boom. It is also contested: its central claim has not held up in the larger reviews that followed. A striking single study is a reason to look harder, not a settled answer.

    Did it hold up?

    The research: Melby-Lervåg and Hulme, 2013

    • In short: Across the controlled studies of working memory training, training reliably improved performance on tasks like the ones practised, but those gains tended not to last and did not carry over to other abilities such as reasoning.
    • What they did: Combined the results of controlled studies of working memory training (a meta-analysis), looking separately at near gains, on similar tasks, and far gains, on other abilities.
    • Who was involved: No new participants. The researchers pooled the results of existing working memory training studies.
    • What they found: Reliable short-term gains on the kind of task people trained on. Those gains tended not to last, and they did not transfer to other skills.
    • What they concluded: Working memory training is not an effective way to improve general cognitive ability.

    A meta-analysis carries more weight than any one study, because it weighs the results of many together. Its conclusion also matches the large review that came three years later, so this is strong evidence. Notice what it does not say, though. The training did improve the trained task. It was not useless. It just did not travel.

    The research: Simons et al, 2016

    • In short: A review of every peer-reviewed intervention study cited by leading brain-training companies found extensive evidence that people improve on the trained tasks, less evidence for closely related tasks, and little for distantly related tasks or everyday thinking.
    • What they did: Examined the quality of those studies and what each showed about transfer: from the trained tasks to related tasks, to unrelated tasks, and to daily life.
    • Who was involved: No new participants. Daniel Simons and six colleagues went through all the peer-reviewed intervention studies the leading brain-training companies cite.
    • What they found: Extensive evidence of gains on the trained tasks; less evidence of gains on closely related tasks; little evidence of gains on distantly related tasks or in everyday cognitive performance. Many of the studies had major shortcomings in design or analysis, and none met all of the best practices the authors set out.
    • What they concluded: Claims that brain-training games improve real-world thinking are not well supported by the evidence the companies themselves cite.

    This is a large critical review rather than a new experiment, published in Psychological Science in the Public Interest. Its strength is where it looked: at the companies' own evidence, examined by a large group of researchers, with a conclusion that sits alongside the 2013 meta-analysis. Strong evidence, with one fair caveat. A review can only judge the studies that exist, and it found those studies often weak.

    Put the three side by side and the picture is consistent. One striking study, then a meta-analysis and a major review that both found the gains stay close to the task. You can say with confidence that brain training improves the trained task. You cannot say it makes you smarter in general, so "brain training raises your IQ" is not something the research supports.

    What I take from it

    Four things stand out to me.

    You get good at what you do. That is the whole finding in a sentence, and it cuts both ways. Give twenty minutes a day to a puzzle and you get better at that puzzle; give them to putting things you need into a memory palace (a set of places you walk through in your mind, in order, each holding a picture of something to remember) and that is the skill you grow. The research does not say training is pointless. It says choose what you train.

    Memory for memory's sake is pointless. That is my view. Memory techniques are thinking tools: you turn information into pictures, place them somewhere you can walk back to, then ask what each picture means. They help you remember so you can think more deeply about what you have learned. The score was never the point. The point is having the information show up in your life when you need it.

    Judge the drill by the skill. Before an app gets your time, I would ask whether practising it is practising the thing you want to get better at; the pack-of-cards test in Memory Games for Adults That Actually Train Your Memory is a quick way to check.

    No magic bullet, either way. Brain games are not a magic bullet, and neither are memory techniques. A memory technique is a skill, so it takes effort, more than rote repetition does. The good news is that anyone can learn it. What I have seen is that using a technique on something real is what lets you feel its value, and feeling the value is what keeps you practising.

    What the research does not show

    These studies get quoted in both directions, so it helps to know where they stop.

    • Not that brain training does nothing. People got better at the trained tasks. That is a real gain, just a narrow one.
    • Not that every study was worthless. The 2016 review found that many had major flaws and that none met every best practice. That is a call for better studies, not proof that all of them were junk.
    • Not that the 2008 result is settled. One study launched a boom; the later reviews mostly did not back its central claim. Quote it as the start of the story, not the end of it.
    • Not anything about memory techniques. These studies looked at working memory drills and commercial brain-training programmes. Memory palaces and other memory techniques have research of their own, set out in Does the Memory Palace Work? What the Evidence Shows.

    Where memory techniques fit

    The memory technique research sits next to the brain-training research in an interesting way.

    In the memory athlete research led by Martin Dresler (memory athletes are people who compete in memory sports), six weeks of memory palace training left beginners with recall that was still well above their starting point four months on. In the 2021 follow-up, the comparison group who did dual n-back working memory drills instead did not get that lasting gain in recall. Both papers are broken down in The Memory Athlete Brain Study (Dresler 2017) Explained.

    There is a classic warning on the other side. In a single case reported by William Chase and Anders Ericsson in 1979, a student known as SF who grouped digits into running times took his digit span (how many digits he could hear once and repeat back) from about 7 to 70 over roughly 200 hours of practice; when he switched to letters, his span fell back to about six. The skill belonged to the material he had practised on. Memorisation Techniques Compared: What Works, What Doesn't, and When covers where chunking fails.

    The SF case says the same as the reviews: practice builds the skill you practise, on the material you practise it on. From what I have seen, a memory technique, once you are proficient, can be pointed at anything: a new list, a new name, a new talk.

    How to use it

    Whether you play brain games or not, here is how to put the research to work.

    1. Name the skill you want. Remembering names, giving a talk without notes, holding on to a revision topic. Be specific: that is the thing to train.
    2. Ask one question of any game or app. Does practising it mean practising the thing you named? If not, enjoy it as a game and expect it to make you good at the game.
    3. Practise on real material. If the goal is remembering real material, practise on it. A word list is a fair way to measure yourself; give the rest of your practice time to what you need.
    4. Start small, today. In your next conversation, pick out three key things the other person says. Turn each into a picture and put it in a temporary memory palace, one you build on the spot, use and let go: three places around the room you are in. Afterwards, walk back through the room in your head and see what is there.
    5. Measure the skill, not the score. Count how many names you still have the morning after the meeting, or how much of the talk you got through without looking down.

    If the memory palace is new to you, How to Build a Memory Palace: A 7-Step Beginner's Guide (With Worked Examples) takes you through it step by step. For games built on the disciplines memory athletes compete in, read Memory Games for Adults That Actually Train Your Memory, and for a regular routine, The Daily Memory Workout. And for the mental workspace the n-back targets, What is Working Memory? covers what it is and why it overloads.

    Memory map

    1. Brain training makes you better at the games you practise. Image: a phone screen full of gold stars, held in a kitchen doorway by someone looking puzzled.
    2. Near transfer is the trained task; far transfer is everything else. Image: a short footbridge over a stream beside a long bridge vanishing into fog.
    3. Jaeggi 2008: dual n-back training seemed to raise fluid intelligence. Image: a set of headphones and a flashing grid of squares on a desk, with a rising thermometer beside them.
    4. The 2008 result is contested, not settled. Image: a trophy sitting on a set of weighing scales that are tipping away from it.
    5. Melby-Lervåg and Hulme 2013: gains on the trained task, short-lived and not carried over. Image: a sandcastle by the waterline, with the tide coming in.
    6. Simons 2016: the companies' own evidence held little for everyday thinking. Image: a stack of glossy brochures under a magnifying glass on a researcher's desk.
    7. Many studies had major flaws; none met every best practice. Image: a row of wobbly chairs, each one missing a leg.
    8. You get good at what you do, so choose what you train. Image: a signpost with one arm pointing at a blank wall and one pointing at a front door.
    9. Memory for memory's sake is pointless. Image: a trophy cabinet full of gleaming cups in an empty, locked room.
    10. The n-back group did not get the lasting gain the memory palace group did. Image: two hikers four months on, one high on a ridge, one back at the car park.
    11. The digit skill did not carry over to letters. Image: a runner's stopwatch full of digits beside a heap of alphabet blocks.
    12. Start small: three key things from a conversation, in a temporary palace. Image: three objects glowing in three corners of a café while two people talk.

    Sources

    • Jaeggi, S. M., Buschkuehl, M., Jonides, J., and Perrig, W. J. (2008). Improving fluid intelligence with training on working memory. PNAS, 105(19), 6829 to 6833. https://doi.org/10.1073/pnas.0801268105
    • Melby-Lervåg, M., and Hulme, C. (2013). Is working memory training effective? A meta-analytic review. Developmental Psychology, 49(2), 270 to 291. https://doi.org/10.1037/a0028228
    • Simons, D. J., Boot, W. R., Charness, N., Gathercole, S. E., Chabris, C. F., Hambrick, D. Z., and Stine-Morrow, E. A. L. (2016). Do "brain-training" programs work? Psychological Science in the Public Interest, 17(3), 103 to 186. https://doi.org/10.1177/1529100616661983
    • Dresler, M., Shirer, W. R., Konrad, B. N., Müller, N. C. J., Wagner, I. C., Fernández, G., Czisch, M., and Greicius, M. D. (2017). Mnemonic training reshapes brain networks to support superior memory. Neuron, 93(5), 1227 to 1235.e6. https://doi.org/10.1016/j.neuron.2017.02.003 (the home for this study is The Memory Athlete Brain Study (Dresler 2017) Explained)
    • Wagner, I. C., Konrad, B. N., Schuster, P., Weisig, S., Repantis, D., Ohla, K., Kühn, S., Fernández, G., Steiger, A., Lamm, C., Czisch, M., and Dresler, M. (2021). Durable memories and efficient neural coding through mnemonic training using the method of loci. Science Advances, 7(10), eabc7606. https://doi.org/10.1126/sciadv.abc7606 (the home for this study is The Memory Athlete Brain Study (Dresler 2017) Explained)
    • Chase, W. G., and Ericsson, K. A. (1979). A mnemonic system for digit span: one year later. Paper presented at the 20th Annual Meeting of the Psychonomic Society, Phoenix, Arizona. https://apps.dtic.mil/sti/pdfs/ADA081039.pdf. Published study: Ericsson, K. A., Chase, W. G., and Faloon, S. (1980). Acquisition of a memory skill. Science, 208, 1181 to 1182.

    Loci

    Build the memory palace in this article inside Loci — store your locations, link images, and rehearse with spaced repetition.

    Try it in Loci

    Memory School community

    Learn directly with Mark inside the Memory School community on Skool — monthly lessons, practice, and Q&A.

    Join Memory School

    About the author

    Grand Master of Memory (1995). Author of three books on memory published by Hodder and Nick Hern. Creator of BBC1's Monkhouse's Memory Masters. Memory coach on Channel 4's Can I Improve My Memory?

    Read full bio →