Neurostimulation - The Genius Machine?

Do you wish you were smarter? Are you often baffled by puzzles?

According to Australian neuroscientists Chi and Snyder, all you need is a bit of electric assistance: Brain stimulation enables the solution of an inherently difficult problem.

In their study, 22 volunteers were faced with the 9 dots problem, a notoriously difficult puzzle. The goal here is to draw exactly four straight lines connecting all nine of these dots, without retracing any line, or lifting your pen from the page.

Can you do it?

If not, don't worry; not many people can. None of Chi and Snyder's 22 subjects did it in the 3 minutes before the stimulation was turned on.

But 5 of the 11 volunteers later managed to do it, after 5 minutes of transcranial direct current stimulation (tDCS), a simple form of neurostimulation in which a weak electric current is passed through the head via electrodes attached to either side. The "L− R+" current was designed to boost the right temporal lobe while inhibiting the left, on the hypothesis that the right side of the brain helps us "think outside the box" (literally.)

None of the 11 volunteers in the placebo control group succeeded. They were given tDCS but after 30 seconds, it was gradually turned off; this is intended to produce the same tingling sensations as real tDCS, but without affecting the brain. That's statistically significant (p=0.018, one tail fisher’s exact test), although the numbers are small.

The authors also refer to other unpublished data:

We would like to emphasize the robustness of our finding. The finding that tDCS enabled more than 40% of participants to solve the ‘unsolvable’ nine-dot problem is consistent with our pilot study (see Section 2), which shows that whereas no one solved the nine-dot problem in the sham stimulation condition, 3 out of 7 participants in the L−R+ stimulation condition did so after stimulation. It is also strongly supported by subsequent studies where we, for curiosity, included the nine-dot problem at the end of an unrelated experiment.

In fact, of all the data we have ever collected by 2 different experimenters over eight months, we found that 0 out of 29 participants in the sham stimulation condition solved the nine-dots problem, whereas 14 out of 33 participants (naïve to the problem) in the L−R+ stimulation condition did so. The probability that by chance 14 out of 33 participants solved the problem is less than 1 in a billion, according to analysis using binominal distribution (assuming that the expected solution rate without stimulation is 5%).

Hmm. When these guys published similar tDCS results with a different puzzle last year, not everyone was convinced. Critics said that 'Thinking caps' are pseudoscience masquerading as neuroscience:

Chi and Snyder's participants solved maths puzzles that the researchers claim required "insight", yet crucially the subjects did not perform any other tasks to show that only puzzles requiring "insight" were influenced by the brain stimulation...

Rather than encouraging novel thinking, maybe brain stimulation made participants less cautious in reaching a decision, or maybe it helped them recall a similar problem seen a few minutes earlier, or maybe it made them temporarily less distractable (or even dulled their hearing), or maybe it boosted general alertness.

The point is that without appropriate experimental controls, the results are virtually meaningless...

Personally I don't think this is all that concerning because the 9 dots problem is really hard and I find it implausible that general alertness would help much; in this study, Chi and Snyder did give people a mental arithmetic task as well to try to control for such non-specific effects. Everyone did it three times - before, during, and after stimulation. However curiously, while they present mental arithmetic data for before and after (showing no effect of tDCS), they don't mention what happened during stimulation. Hmm.

More worrying is that although the study was placebo controlled, the authors don't say whether subjects were randomly assigned to active or placebo tDCS; if not, that's a major flaw. And although the subjects weren't told which group they were in, at least one of the experimenters must have known because someone was manually controlling the tDCS current switch. Were they in the same room as the subject? Could they have, unconsciously, been influencing them?

I'll be skeptical until we get some independent replication.

Here's the 9 dots solution, for those of you who didn't have a tDCS machine handy...

Chi, R., and Snyder, A. (2012). Brain stimulation enables the solution of an inherently difficult problem Neuroscience Letters DOI: 10.1016/j.neulet.2012.03.012