What's the best way to test cognitive function, inexpensively, privately, and repeatably?

March 6, 2009 8:20 AM

What is the best way to test cognitive function, inexpensively, privately, and repeatably? I'm looking for some semi-objective means of evaluating different tweakings of my head-med cocktail for Adult ADD and Depression. (I am experimenting under the care and guidance of a physician and am reasonably well versed in the pharmaco-kinetics and -dynamics of my meds -- so please note that I'm not soliciting advice concerning the dangers of making one's own medication decisions.)
posted by anonymous to health & fitness (10 comments total)

I know one dude who used the game Set here - he looked at how many sets he could make in a given time frame, and used that as a very crude benchmark.

Huge problem: I think one can get better-skilled at the game quickly, especially at first.
posted by amtho at 8:32 AM on March 6

I think one problem (as amtho points out) with the simple DIY testing you can do, is that you'd tend to get things which you'd get better at over time. The game Set would be one such thing, but also other things like drawing a complex figure from memory, or those find the next image in this sequence you find on all those onilne IQ tests are things which you will get better at if you do them more often.

You'd need to measure a number of times for each "cocktail" since normal daily variances play into it as well, so you really need something which you normally wouldn't get better at even if you do it quite a lot (or which you max out your proficiency at quickly so once you've learned something you stay on the same level).


Having said that, perhaps buy a few puzzles with a large number of pieces (>500) and see how long it takes to finish them? Puzzle laying is something I'm sure you get better at, but you could probably reuse puzzles for each cocktail since you won't be able to completely memorize every single piece and where it goes (I know I wouldn't be able to at least).
posted by bjrn at 8:39 AM on March 6

Addressing amtho's very good point, you may wish to find a test at which you cannot become much better after hitting a plateau.

I'd suggest something like Mastermind. While there are (complex, memory intensive) algorithms that lead to about a four move average game, feasible human algorithms tend toward six moves. In actually play, because each move's answer must be reconciled with previous answers to eliminate possibilities, I find it a useful gauge of concentration/tiredness. Because the best algorithm (for me, givven my short-term memory and other predilections) has reached fixation and does not change/improve, that gauge isn't "stretched" or "compacted" as amtho warns.
posted by orthogonality at 9:22 AM on March 6

ADHD is usually tested clinically with either a CPT (Continuous Performance Test) or TOVA (Test Of Variables of Attention). It's quite common to use these when adjusting medication types and dosages. [Interestingly, with these tests, people do not get better over time, they actually get worse a bit, then plateau off.]

If you're a programmer, you could probably write one of your own you can use for free. You'd have only yourself as a baseline, but that might not be too bad.
posted by adipocere at 9:25 AM on March 6

I sorta kinda do this with sudoku. I use the site created by a MeFite, which gives me consistent results.
posted by Goofyy at 9:36 AM on March 6

The Stroop Test! There's a lot of them out there, but this one is interactive.
posted by fiercekitten at 9:43 AM on March 6 [2 favorites]

I'd say that something to consider is what part of you're ADD you're most interested in monitoring. Is it sustaining attention over time? Is it resisting distraction/interfering information? The type of task you choose should be tailored to what you're hoping to improve.

Memory span ("simon") tasks are commonly used to assess ADD symptoms, as are tasks like the symbol search or coding (from the WAIS). You could create your own versions of any of these, and they should be relatively resistant to practice effects, if you change the stimuli. I would think that your doctor (a neuropsychologist?) would be able to recommend some other relevant tasks.
posted by supramarginal at 5:49 PM on March 6

Speaking as someone who is ADD - I'm not sure I buy into any of these as being all that useful of a test for ADD. I can, for example, do very well at the "Simon" thing. Where I die a horrible death is trying to do something like that and something else. In computer terms I'd say I can handle a very deep stack, but I can't maintain more than a few of them.

When I first realized how different I was from "normal" was when there was some construction going that was noisy as all getout and making it impossible for me to get anything done. Most of my co-workers were like, "Oh, I just ignore that," which, to me, seemed like suggesting that I try not seeing the color blue.
posted by Kid Charlemagne at 6:31 PM on March 6

The Stroop Test! There's a lot of them out there, but this one is interactive.

Yes, the Stroop test is relatively unaffected by learning and practice effects, however it only really tests one's level of concentration/reaction time, and this isn't the full gamut of cognitive ability at all, though it is somewhat of a predictor, I suppose.
posted by tybeet at 12:55 PM on March 7

This guy blogs about self-experimentation and brain function. You might find some worthwhile ideas there.
posted by DarkForest at 8:15 PM on March 9