Yeah, I’ve been very busy and haven’t posted our results yet, so I might as well post preliminary ones here.
We had five people doing Dual N-Back; everyone did it for ~40 minutes to an hour in three sessions. The control group just did three sessions; the experimental group did a session, applied the electrodes, did a session, removed the electrodes, and then did a final session.
The control group showed minor improvement from session to session. Both experimental subjects showed a minor decline under current application which reversed after the current was removed. The other guy didn’t notice much originally, I reported that I felt less interested in doing DNB under current, and found the data suggested my ability declined; he analyzed his data and agreed it was probably a negative effect.
Our equipment was poor, though. I’m pretty sure we should have used EEG electrodes, and instead used TENS electrodes (used for muscle stimulation) which are way bigger- which probably had some effect. We were working off a 10-20 diagram, but none of us are trained technicians for this sort of thing.
So, the takeaway for me was that this does have the power to change things, but getting that change to be positive is nontrivial. I’m planning on investing more in equipment (we can get proper electrodes and a placement cap for $70, whereas the electrodes I bought were around $4 a person) and repeating (with me, at least; I don’t know how many other people are willing to shell that out). Once I do that, I’ll write that up.
Also, doing a sniper game might use different mental faculties than dual n back.
Right. tDCS has a number of parameters you can vary, of which the most obvious are location of electrodes and polarity of electrodes. We were trying to stimulate the dorsolateral prefrontal cortex, done by running a current across two scalp locations; the article mentions temple and arm placements that I don’t think I saw elsewhere in the literature (but have the equipment to try that out, so I think I shall!).
The polarity is the other huge thing- if cathodal stimulation lowers the polarization and that helps, then anodal stimulation should raise the polarization and that should hurt. (One of the early proposed therapeutic uses was putting the brakes on brain areas that were overactive in negative ways, like anxiety regions.) It’s very possible that we screwed up and misread the paper and applied the voltage in the wrong way, and if we swapped it around it would help.
Shockingly, it didn’t seem to make much of a difference in improving n-back results. That seems to be the current consensus (I don’t think the others would put up much resistance to my saying this.)
Previous post where tDCS was discussed; there’s also a dedicated subreddit.
Seems the crazy LessWrongers in Austin planned to try it, but I don’t know the result.
Yeah, I’ve been very busy and haven’t posted our results yet, so I might as well post preliminary ones here.
We had five people doing Dual N-Back; everyone did it for ~40 minutes to an hour in three sessions. The control group just did three sessions; the experimental group did a session, applied the electrodes, did a session, removed the electrodes, and then did a final session.
The control group showed minor improvement from session to session. Both experimental subjects showed a minor decline under current application which reversed after the current was removed. The other guy didn’t notice much originally, I reported that I felt less interested in doing DNB under current, and found the data suggested my ability declined; he analyzed his data and agreed it was probably a negative effect.
Our equipment was poor, though. I’m pretty sure we should have used EEG electrodes, and instead used TENS electrodes (used for muscle stimulation) which are way bigger- which probably had some effect. We were working off a 10-20 diagram, but none of us are trained technicians for this sort of thing.
So, the takeaway for me was that this does have the power to change things, but getting that change to be positive is nontrivial. I’m planning on investing more in equipment (we can get proper electrodes and a placement cap for $70, whereas the electrodes I bought were around $4 a person) and repeating (with me, at least; I don’t know how many other people are willing to shell that out). Once I do that, I’ll write that up.
Some posible problems:
what type of device did you used ?
current?
current density of electrodes?
refference electrode position?
use EEG or TENS electrodes is bad idea they have wrong current density and resistence and dnagerous electrochemial products
use saline soaked electrodes of right current density
http://brmlab.cz/project/brain_hacking/tdcs http://www.longecity.org/forum/topic/50076-tdcs-thread/page__st__30__gopid__503859#entry503859
Also, doing a sniper game might use different mental faculties than dual n back.
That particular tTCS might or might not work as well if the people who you were supposed to shoot or not shoot changed according to complex rules.
Right. tDCS has a number of parameters you can vary, of which the most obvious are location of electrodes and polarity of electrodes. We were trying to stimulate the dorsolateral prefrontal cortex, done by running a current across two scalp locations; the article mentions temple and arm placements that I don’t think I saw elsewhere in the literature (but have the equipment to try that out, so I think I shall!).
The polarity is the other huge thing- if cathodal stimulation lowers the polarization and that helps, then anodal stimulation should raise the polarization and that should hurt. (One of the early proposed therapeutic uses was putting the brakes on brain areas that were overactive in negative ways, like anxiety regions.) It’s very possible that we screwed up and misread the paper and applied the voltage in the wrong way, and if we swapped it around it would help.
Shockingly, it didn’t seem to make much of a difference in improving n-back results. That seems to be the current consensus (I don’t think the others would put up much resistance to my saying this.)
There was also a very detailed comment thread on Hacker News—someone claimed to have built one from 30$ in parts.