If you just want something that appears to be subconcious, clenching your jaw can produce quite a distinguishable movement while not appearing very obvious to an outside observer.
[nitpick]
An MRI cannot, but a PET scanner can.
[/nitpick]
Even more of a nitpick, an MRI can. It’s called functional MRI and it’s what I do every day. Instead of a single image it takes a sequence of fast images. It works by tracking the oxygenation of the blood cells. There’s some debate over whether or not this indirect method is precisely accurate, but for something like this it would work fine. With the exception of cost, size, energy requirements, etc.
And for somebody who wanted to know how much they cost, I think a general rule of thumb is about $1 million per tesla.
Oops, I forgot to add, not to mention the cost of installing and making sure your scanner room is properly shielded and such.
Well I’ll be dipped. You learn something new every day.
Not to hijack any further, but there was an interesting article in nytimes maybe? about a researcher who was able to tell if a subject was looking at vertical bars or horizontal bars, just from the fMRI images. Amazing.
When I was in middle school I was thrown against a cinder-block wall and passed out. They later ran me through an electro-encephalogram (EEG), which consisted of putting a bunch of electrodes on my head. Didn’t even have to shave my hair off.
Too bad you’re not in chem. After finishing the AP and the seniors left, the underclass among us made things go boom.
Come on, how many science classes can commandeer the football field to set off thermite?
and
You don’t have just 10 hours, you have your ten, plus ten from each of any other students you can convince to help. I’d suspect at least some of the other students would think this was pretty cool also. And possibly the teacher as well.
Even with your 10 hours and everybody else’s 10 hours, it’s going to be quite a challenge. If you’re going to go the EEG route, then you’ll have to figure out where to place the sensors. Then remember that even eye blinks can create a signal that will swamp out your brain activity. In the ERP studies we do, we have to have the person sit as still as possible. Moving around is going to cause even greater signal noise.
EEG signals are on the order of microvolts so you’ll need a pretty good amplifier. The training is going to be difficult as well since it won’t be immediately obvious how to send out the correct signals. Given a semester I think you could do it, but with only 10 hours… I like to think that people can do anything, but your time constraints are going to make this very, very difficult.
Maybe they can get it built in the time allotted, then he can spend all summer trying to get the colors to change. Next fall, he’ll get the “What did you do last summer” writing assignment. Ummm, I made a light turn on. 
You know, radios catch the modulation of carrier frequencies at pretty low signal strength, and turn it into audio. You’d have the tuning for frequency and the output all built in, and in a small radio it might all be on a small cheap card.
And radios can turn on a light to show when you are on vs. near the carrier. If that ability isn’t built into the cheap card, you could get two cards. Tune them to different “carriers”, ie alpha vs beta waves, or whatever you find by experiment can come from your skull, such that in each case the card output would turn on an LED rather than running a speaker.
Here’s hoping someone who actually has knowledge might tell us if this could work.
Get the thermometer from the medicine cabinet and check your temperature. Quickly!
20 years ago, when I was 8, I bought a “Lie Detector Kit” from Radio Shack, for $6.00.
It was basically a biofeedback meter with maybe 20 parts, mostly semiconuctors.
It read like a Geiger counter, in that with the pads on my index finger, I could “think hard” or stress myself with thought, and the clicks would speed up to a tone.
Surely making an LED light up from a given threshold doesn’t require assembly langquage.
errrrr 25
A relatively cheap opensource EEG,
circuit here
but still not cheap.
NOTE: I didnt check if the above circuit includes one, but it is a good idea to put an optoisolator between the probes and the machine so you dont put a nasty current into your head.