Study finds Lumosity users show no improvements in cognitive performance
arstechnica.co.uk
arstechnica.co.uk
A Luminosity recruiter contacted me after I left, and I asked for evidence that their research is actually helpful. They were very kind and passed along a number of studies that they'd done. I wasn't super impressed. Among other things, they were marketing their games to the general public, not the populations represented in their small studies. Quite honestly, aerobic exercise causes a huge upregulation of neurogenesis (and many other positive things), so people would be better off going for a light jog 2-3 times a week. It felt like they were taking advantage of families desperate to help a family member with dementia and willing to try anything.
Aside from the diet/exercise/socialization/meditation, what should people be aware of?
Are there some effective supplements/nootropics, or some promising research to look into?
Regular aerobic exercise (stuff that elevates your heart rate) is still the best thing we know of. It's good for upregulation of neurogenesis and BDNF.
There are dozens of papers out there showing that a certain treatment/activity/drug increases neurogenesis. It's funny, I was talking about neuroscience to a lyft driver a way back who was curious about neurogenesis/neuroplasticity, and I told him that exercise is your best bet, but we also had some good results treating our brain-damaged rats with an acetylcholinesterase inhibitor. He shrugged off the exercise bit but of course asked me to spell out ACh inhibitor for him (which I really wouldn't recommend any random person off the street taking).
I experimented for a bit with nootropics after I graduated. Gwern's posts are really interesting if you haven't read him (https://www.gwern.net/Nootropics). A number of people have found success with a basic choline and -racetam stack. As much as tech blogs love blogging about how we're all hooked on nootropics in the valley, it's unlikely that taking some pills in the morning is going to dramatically change your life (though it might nudge it in the right direction). As one of the other replies said, lifestyle changes are still the most effective approach.
Here's an article where he compares it to cold fusion: http://www.nytimes.com/2012/04/22/magazine/can-you-make-your...
She has long stated that Lumosity is crap. For two three main reasons:
1. The "improvement" you see in playing Lumosity games is due to getting good at the game, not actually underlying improvements.
2. The games you do want to pursue need to to be strategic and challenge the mind. They need to be different, and involve complex scenarios. She likes Rise of Nations, by way of example. Flash minigames don't cut it (and neither do crosswords).
3. Brain games are 20% of a total set of activities that have an impact on brain health, the other 80% including physical exercise, diet, socialization, and lack of stress (mindfulness training helps here).
Anyway, fascinating stuff, and glad Lumosity's "bad science" and false claims are coming to the fore.
If you'd like to improve your memory I suspect you'd be far better off studying the masters from Aristotle to Dominic O'Brien whose memory palace techniques seem incontrovertibly proven if you consider the techniques of the winners of the world memory championships. [2] [3]
Finally, meditation is not causally linked to higher levels of grey matter but the research seems promising [4]
(I tried to find decent links but being at work can't spend a significant portion of time on the post, these should be good starting points for anyone interested).
[0] - http://www.health.harvard.edu/blog/regular-exercise-changes-... [1] - https://www.livescience.com/46048-learning-new-language-brai... (sorry for the ads, and I think it autoplays a video :( ) [2] - https://en.wikipedia.org/wiki/Art_of_memory [3] - https://en.wikipedia.org/wiki/Dominic_O%27Brien [4] - https://www.sciencedaily.com/releases/2015/02/150205142951.h...
"The researchers note limitations in the study, most importantly that they only looked at young, healthy people. Brain-training games may still prove useful for aging adults or people suffering from addiction or unhealthy behaviors, the researchers caution."
[1] we only studied marathon runners"
That study of two mile walks for marathoners would suddenly be worthwhile if a popular, well-funded program claimed they could improve your marathon time with a two mile walk.
This study shows that there is no basis for the claim that you can enhance performance of normal/healthy performers.
The researchers state the have insufficient evidence to determine the effect on low/unhealthy performers.
This study is about people who've never used Luminosity's techniques, but are in unrelated good (brain) health. There are a lot of known ways to mitigate or undo damage to the brain. Luminosity made a much bolder claim, that they could raise IQ in healthy, high-functioning adult brains. That's the Holy Grail of intelligence research, because no one has ever found a way to do it.
It's like claiming a two mile walk each day will make literally anyone a much better runner. If you claim that, it's relevant that your offering is no good to marathoners.
Given the context of the article, this is an unscientific and immoral statement.
It is irrational to believe something is true or possible unless you have evidence for it.
If the researchers have a positive reason to believe brain-training games are a form of medical treatment for aging or addicted populations, they should state that. None of this "may" nonsense.
This is actually two statements.
> It is irrational to believe something is true unless you have evidence for it.
Sure, I'd agree with that.
> It is irrational to believe something is possible unless you have evidence for it.
Whoa, there. This seems like it goes...way beyond rationality. I would argue that you would need strong evidence if you were going to claim something was impossible.
If you have no evidence for a claim, then any statement made regarding that claim, either in support of or opposed to, is arbitrary and lacks any intellectual value.
For example, any claim about the (non)existence of god is arbitrary and not worth even a moment of consideration, because there is no evidence whatever in support of the claim either way.
There is a hard claim of possibility, that is a claim of the state of nature (e.g. it is possible for a photon to travel at 0.5c). This hard claim is one where we definitively say that it is _certain_ that it's a physical possibility.
Then there is a soft claim of possibility, that is a claim of our state of knowledge. (e.g. It's possible that Rob went to the store, but I'm not sure).
These are two different claims. One is a hard claim that should only be made with evidence. The other is a statement of our uncertainty and an open mind; that doesn't need any evidence to make.
It's a shame that these two different meanings of these words exist, but such is the English language. They are saying that their research does not close the door on that eventuality, and that their state of knowledge on that topic is uncertain.
If, instead, you know Rob well and you happen to have a general understanding of his schedule and needs, you might respond with "It's possible he went to the store; last night he mentioned he ran out of toilet paper." This is an evidence-backed claim.
A claim about the state of nature is the same. Your knowledge leads you toward certain avenues of research, so you might think, "photons have properties x, y, z, so maybe they can travel at .5c, let me look into it." However, it would be improper to base your research on whim, like "I think photons are little gremlins and they run really fast, but can they run slow too?" Such a claim is not a claim about reality, it is arbitrary.
Claims about the world must be grounded in knowledge, and knowledge must be based on evidence.
No, the claim about what is possible for a photon was much stronger. I am asserting now that photons can travel at .5c, full stop. I will not look into it, as I have already determined it is possible. I am not making a hypothesis about the universe, I am making a claim about how I believe it to actually work. Anyway, that is an aside.
I still think this just comes down to semantics about what is a claim about the world and what is a statement about uncertainty.
I interpret them saying something is possible as them making a statement about their own partial knowledge. I specifically do not interpret it as a claim about the world. I do interpret it as a claim about their own research and their own partial knowledge.
Their research does not demonstrate an impossibility. That is the claim they present. I still believe you are making a semantic interpretation of the word "possibility" that is different from the intent of the researchers. I think you are interpreting a stronger claim than the researchers are making.
Here is, for example, the same qualification as reported in a different article:
> Although, in this study, the researchers found that commercial cognitive training alone would not have an influence on one's decision-making process or cognitive abilities, they believe that it was still an avenue worthy of rigorous investigation.
https://www.sciencedaily.com/releases/2017/07/170710135410.h...
If you have no evidence against it, it is not irrational to believe something is possible.
If I claim there are tiny, sentient gnomes running around in my garden, I should have some evidence for that claim. If I insist it's possible without identifying any evidence, you would rightly conclude I'm insane.
You evaluation of evidence, or your logical process, could both be mistaken and so you might believe things that are not true. But if your belief is based on evidence, you can test it and revise your understanding.
It is neither unscientific nor immoral to note a popular belief which the study does not provide a basis for rejecting on which a conclusion might be likely to be improperly formed based on the study.
"May" is not a statement of positive belief.
"May" is a word that does not belong in scientific language. If there is evidence for a claim, you state it. If there evidence against it, you state it. If there is no evidence either way, you say "I don't know."
"May" implies something is possible without the burden of identifying the evidence that suggests that possibility.
Consider a different statement: assume there was a study that showed rhino horn is not effective in improving mental performance in young adults, and then you read "Rhino horn may still prove useful for those suffering from cancer, the researchers cautioned." How would you react?
I'm debating just setting up a spaced repetition app with my own custom knowledge base. At least that way I may not be getting higher brain function, but higher retention on memory items I care about.
1.Learn a Second Language 2. https://projecteuler.net/archives 3. http://rosalind.info/problems/list-view/
If you are going to exercise your brain, you might as well get smarter and learn useful skills in the process.
One might argue that that's the only way to actually exercise your brain.
The notion that a startup has A: deep insights into the nature of cognitive function that others do not; and B: devised and tested a paradigm that sustainably improves it, was always a little shocking to see claimed.
Importantly, Semmelweis couldn't offer an explanation for why washing hands helped, which seems to have greatly hindered his credibility. I said you were offering a germ-theoretic perspective, because we do (and must do) the same thing: How else should we separate the spurious correlations from the important ones? How do we separate superstition from good practice? How do we tell the crackpots from the creative? If we suspend the germ theory, what if there was something else going on at the hospital, and hand-washing was just something this guy noticed and latched on to? Then you'd be giving people the wrong advice---which is, after all, one of many reasons that correlation is not causation.
As strange as it sounds though, I think this has been making me more... dumb. Sure, I'm improving a subset of skills, especially computer science related with some of the more challenging projects, but I don't feel more intelligent. I forget things constantly, and frankly my knowledge of the world seems to just.. fade. I'm becoming so hyper focused that it's hurting me.
That's why I mentioned spaced repetition. Not only do I want to improve my general brain functions, but I especially want to improve my recall of specific things. So I'm using a distributed database of mine as a way to keep a life journal. Anything I forget, I want to add to this knowledge journal, and try to ensure that I keep things in my mind.
Dual n-back was thought to be the only thing that worked, but even that now looks non-transferrable. (Gwern is great on this topic.) For young, healthy adults, I do not know of any brain exercise capable of directly raising IQ or its analogues.
You might, maybe, be able to boost neurogenesis. Exercise is a decent candidate for that, and I don't know the situation with drugs there. But I don't know of games that do it.
Realistically, you've got a few options. Your idea, spaced repetition of useful stuff, is great. Practicing memory tricks like mind palaces can improve your long-term memory on arbitrary facts. Picking up new topics dissimilar to the ones you know probably helps synthesis. Lots of things can maybe delay dementia. Meditation can improve focus and alertness.
All of those are worthwhile. The top-line, raise basic IQ stuff? To my knowledge doesn't exist.
That is, does it make it easier to recall things that are not part of the spaced repetition training?
You can improve your performance on almost any single task, which is great news. Certain forms of learning (e.g. language) help defer dementia, exercise and meditation improve focus, memory tricks improve recall, learning diverse topics improves synthesis.
But in terms of 'intelligence' in the clinical sense? Despite many people's efforts, I've never seen a single convincing success for healthy adults increasing their IQ. You can practice a specific test until it's inapplicable, but if you take one test, try to raise your IQ, and take another test, you won't see gains. Dual N-back is still the best thing known, and it doesn't work appreciably.
Fortunately, the practical stuff - memory, focus, breadth of knowledge - is awfully helpful in real life. But I've never seen any evidence at all that intelligence in the IQ sense can be raised in adults.
No doubt people have different ceilings for their potential, but I feel like most people who are viewed as smart arent just leaning on natural talent, rather they've cultivated skills that get the most out of what they have
And I agree on the ambiguity around 'innate ability'. One could argue that general health is 'subtractive', that bad sleep/diet/exercise are drawing down potential that would exist otherwise. But that's far less clear-cut than something like "only works in dementia patients" or "only helps with Korsakoff Syndrome". I think there's something to it - lifting heavier weights and sleeping more are not unbounded sources of improvement - but I think it's also kind of a copout. If those things operate on a continuum, and people we consider "healthy adults" can still see benefits from them, that's what we wanted to find.
What one can actually accomplish is what matters most. If one is able to improve the layer above raw IQ, the fundamental mental models that provide the base for interpreting, analyzing, and synthesizing everything that happens to them, that would provide immense benefits.
It seems to me that is possible, to an extent. This is something I strive for, and I think I have made a bit of encouraging progress. Of course, evaluating this is... uh... not quite straight-forward, to say the least :)
forgive me if i don't take this meta-analysis seriously.
Critically if you actually read the article you can judge for yourself whether they know what they're talking about, as it's going to be both incredibly in depth and include disclosures of basically all caveats.
>Pre- and post-training, participants completed cognitive assessments and functional magnetic resonance imaging (fMRI) during performance of validated decision-making tasks: delay discounting (choices between smaller rewards now vs. larger rewards in the future) and risk sensitivity (choices between larger riskier rewards vs. smaller certain rewards).
You could certainly imagine doing things to get better at these kinds of tasks---including practicing these tasks. Whether this gets at some kind of murky "cognitive performance" thing depends on how much you're into phrenology and dualism.
Since it seems like you want a scientific answer to this question, you'll always have to confront the realities of experimental design, and because of that, you'll have to nail down exactly what it is you're after. And "cognitive performance" isn't nearly specific enough.
Dual n-back, the general benchmark for these things, is hard. It's exhausting like intricate coding or a tough math proof is exhausting. Certainly I'd be suspicious of anything claiming to do better than that with less mental effort.