For the last four years or so I have daily measured how well my brain is working by means of balance measurements and mental tests. For three years I have used a test of simple arithmetic (e.g, 7 * 8, 2 + 5). I try to answer as fast as possible. I take faster answers to indicate a better-functioning brain.
Yesterday my score was much better than usual. This shows what happened.
My usual average is about 550 msec or more; my score yesterday was 525 msec. An unexplained improvement of 25 msec.
What caused the improvement? I came up with a list of ways that yesterday was much different than usual, that is, was an outlier in other ways. These are possible causes. From more to less plausible:
1. I had 33 g extra flaxseed last night. (By mistake. I’m not sure about this.)
2. The test came at the perfect time after I had my afternoon yogurt with 33 g flaxseed. When I took flaxseed oil (now I eat ground flaxseed), it was clear that there was a short-term improvement for a few hours.
3. Many afternoons I eat 33 g ground flaxseed with yogurt. Yesterday I ground the afternoon flaxseed an unusually long time, making made the omega-3 more digestible.
4. I did kettlebells swings and a kettlebell walk about 2 hours before the test. These exercises are not new but usually I do them on different days. Yesterday was the first time I’ve done them on the same day. I’m sure ordinary walking improves performance for perhaps 30 minutes after I stop walking.
5. I had duck and miso soup a half-hour before the test. Almost never eat this.
6. I had a fermented egg (“thousand-year-old egg”) at noon. I rarely eat them.
7. I had peanuts with my yogurt and ground flaxseed. Peanuts alone seem to have no effect. Perhaps something in the peanuts improves digestion of the omega-3 in the flaxseed.
8. I started watching faces at 7 am that morning instead of 6:30 am or earlier.
Here are eight ideas to test. Perhaps one or two will turn out to be important. Perhaps none will.
After I made this list, I read student papers. The assignment was to comment on a research article. One of the articles was about the effect of holding a warm versus cold coffee cup. Holding a warm coffee cup makes you act “warmer,” said the article. Commenting on this, a student said she thought it was ridiculous until she remembered going to the barber. She sees the person who washes her hair (in warm water) as friendly, the barber as cold. Maybe this is due to the warm water used to wash her hair, she noted. This made me realize another unusual feature of yesterday: I had washed my hair in warm water longer than usual. I think I did it at least 30 minutes before the arithmetic test but I’m not sure. In any case, here is another idea to test. I found earlier that cold showers slowed down my arithmetic speed.
This illustrates a big advantage of personal science (science done for personal gain) over professional science (science done because it’s your job): The random variation in my life may suggest plausible new ideas. As far as I can tell, professional scientists have learned almost nothing about practical ways to make your brain work better. You can find many lists of “brain food” on the internet. Inevitably the evidence is weak. I’d be surprised if any of them helped more than a tiny amount (in my test, a few msec). The real brain foods, in my experience, are butter and omega-3. Perhaps my tests will merely confirm the value of omega-3 (Explanations 1-3). But perhaps not (Explanations 4-8 and head heating).
Butter in specific, or cream in general?
Seth, I have a question about your test. I think that the one at Genomera is identical to the one you use, correct?
If so, my understanding is that the test does not count incorrect answers when calculating average speed. So if you answer 8 questions correctly before you answer 32 questions right, you use only the response time for those 32 questions to determine your brain function.
But what if a faster speed sometimes indicates poorer brain functioning? For example, let’s say that you’re getting many “hard” questions (e.g., “13-7″) wrong, so your results are biased towards easy questions (e.g., “8-8″). Even if your brain is functioning poorly, you will be very fast at those easy questions. So because of this sampling bias, your test may actually give backwards results in some cases.
I also use the genomera test, so I’m curious about your thoughts on this.
I use the 20-question quiz (addition, subtraction, and multiplication) from the Brain Tuner Lite iPhone app. The quiz involves pressing either a “right” button or a separate “wrong” button to say whether a given simple math equation is wrong or right. Since the app adds five second penalties for each wrong answer, it would seem to avoid the problem @Thomas describes.
I perform several tests in succession to measure my cognitive functioning so that I can avoid the problem of an unusually easy batch of questions suggesting improved cognitive function. For example, following the addition of bulletproof coffee to my diet and 20 minutes of dual n-back training to my daily routine, I decreased my speed from an average of just over 600 msec/problem to an average of under 575 msec/problem. Interestingly enough, this increase happened notwithstanding 30 minutes daily of icepacks on the back of my neck and on my trapezis prior to bed (for weight loss purposes as described by Tim Ferriss).
Ajb, I have not varied the amount of cream I consume. I drink a little bit of cream during the day. I suppose butter and cream will have the same effect, but I have not tested that idea. At first I would use only one brand of butter but now I am sure that many brands produce the benefit.
Thomas, the Genomera test is very similar to mine, yes. And, yes, I count only correct answers. About the possibility of bias: 1. Variation in error rates is a problem, yes. But I have found wide variation in error rate to be uncorrelated with average speed. There is no clear speed-accuracy tradeoff. 2. I correct for difficulty. Before averaging across trials, the score on every trial (in msec) is adjusted by the difficulty of that trial. Suppose the trial was 7+3. If 7+3 is on average 50 msec slower than average, the score on a 7+3 trial has 50 msec subtracted before I average over trials. IBecause of this adjustment, it is no help to get the hard questions wrong so that your score will be based only on easy questions.
I would have thought that this was within the normal variation for this kind of test. Some factors that maybe you didn’t think of/account for: hours of sleep the previous night, time the test was taken (in hours after waking, and hour of day), time spent doing analytical things (research, writing, reading) the day before, time spent socializing the day before.
Also, do you have a link for the software to run this test? I’d be interested in tracking my own results.
Seth, off-topic, but along the lines of one of your interests:
“The New Einsteins Will Be Scientists Who Share
From cancer to cosmology, researchers could race ahead by working together—online and in the open”
https://online.wsj.com/article/SB10001424052970204644504576653573191370088.html?mod=googlenews_wsj
O/T but I thought that this might interest you.
https://www.dutchnews.nl/news/archives/2011/10/tilburg_professor_faked_data_i.php
(H/T Naked Capitalism)
Two cariations on flaxseed and yogurt
1) i noticed that the ground seeds are dry when mixed with yogurt. I usually out them first in a little water, and then add yogurt.
2) if left in water for ten or five minutes, it gets a whole different texture and feels different. May chAnge avilability and digestion
Now, I noticed that long grinding makes the ground very hot. With probably effects.
Ps. When i eat flaxseed and yogurt at night i sleep very. Well. Such a sweet wakeup that I rarely had.