Man, this thing is literally what I have been working on since last week. Now it's mentioned in NS I have to rush like there's no tomorrow.
The n-back test is basically a short term memory task. It is has been found to be a strong indicator of fluid intelligence because it is determined by attention and ability to manipulate and perform comparisons on objects stored in the STM buffer. 2-back test is fairly easy without distractions; 3-back becomes very taxing for the untrained. And it can definitely be improved.
Actually I wrote my first n-back test close to two years ago, but for myself. The biggest reason why I wanted to make a web version is because of all these brain games popping up, made by people who don't know what they're talking about. This thing is for real.
I'm also working on subitization test variants, but that's much more familiar.
Haha, back then I wasn't really into database applications and built it just to experience it. That's how I know that 2-back is fairly easy and 3-back isn't, but beyond that I can't answer you (yet).
I gave my dad a 3-back test, on which he did surprisingly well, and he said, "oh, that's easy, all you have to do is concentrate." Indeed, a big portion of it is attention.
A simple n-back variant is live as of this morning but my submission didn't last long... I'm replying to all the responses in this thread. Again, it's a simple implementation, but FYI, http://cognitivefun.net/test/4
Good point. I also don't think anybody would play brain games for very long at a time. But then, such apps are highly motivation-dependent. There is a thing about learning rate and conscious effort, which is a much more difficult problem.
Won't be "done" in a while, because it's part of a larger work in progress, but I should be able to show something soon. I just started learning ActionScript, but cogsci programs aren't hard to write... but who knows, I look forward to criticism, which I expect much.
Hey thanks, it's probably real easy for you to implement, but if you don't feel like writing your own, here's the one I was working on: http://cognitivefun.net/test/4
I submitted it here but it didn't last and I don't want to spam :)
I thought of this question before, and I don't think so.
I think a boosted IQ would precede boosted math problem solving ability, because it deals very much with abstractions and an efficient working memory: "suppose x = a, y = c, z = g, then if xyzabc123, qwerty." Each additional step puts increasing stress on the WM buffer since everything needs to be maintained simultaneously for I/O.
You can, however, improve your speed at solving well-defined math problems by practicing them a lot. I believe this would eventually utilize episodic memory (HD space) rather than WM, which uses a different mechanism. But the speed would probably not generalize to novel problems, which is what I think you're getting at.
The n-back test is basically a short term memory task. It is has been found to be a strong indicator of fluid intelligence because it is determined by attention and ability to manipulate and perform comparisons on objects stored in the STM buffer. 2-back test is fairly easy without distractions; 3-back becomes very taxing for the untrained. And it can definitely be improved.