Jimmy sent this along:
Still, Mr. Perry wondered whether caffeine would help him. When he retired from rowing last July, he decided to do a randomized, blinded, placebo-controlled experiment on himself.
He noticed that the 200-milligram caffeine pills look exactly like vitamin C pills, allowing him to code the pills so that he would not know which one he had taken. For eight months he tested himself once a week, taking two pills an hour before working out on a rowing machine. Then he worked as hard as he could for an hour, recording the results, also recording his guess about whether the pills he took contained caffeine. Mr. Perry, who also is a runner, said that an hour on the rowing machine is the equivalent of an hour of very fast running on the road.
When he finished his study and broke the code late last month, he was astonished to see how much the caffeine had affected him. He was stronger — his power output was 3 percent greater — and faster. In fact, he said the average speed for his tests when he used caffeine was faster than his fastest speed when he was not using caffeine.
He also guessed right most of the time about whether the pills he took were caffeine or vitamin C. Mr. Perry said he is now sorry that he never used caffeine when he was competing. “It would have been a pretty harmless way to do better,” he said.
Yes, neat, but… "Exercise physiologists have studied caffeine’s effects in nearly every iteration: Does it help sprinters? Marathon runners? Cyclists? Rowers? Swimmers? Athletes whose sports involve stopping and starting like tennis players? The answers are yes and yes and yes and yes." So how much is really added here?
How was he able to code the pills himself, than take the pills without knowing what he was taking, and then decode things and figure it out what he was taking later?
Some variation of the following would work. For a two week experiment, take 14 identical Dixie Cups and write the numbers 1 to 14 on the bottom. Put caffeine tablets in cups 1 through 7 and placebos in cups 8 through 14. Place all cups on a table and mix them up thoroughly. Then, selecting cups "at random" (and making sure not to look at the numbers on the bottom) line them up in one long row.
On day 1 take the pill out of the first cup. On each succeeding day, take the next pill in line. When the experiment is completed, turn over the cups to see what the number was for each day.
This is nice and creative! Is there any data so maybe some tests can be done instead of descriptive results?
How the heck can you do a double blind test on caffeine pills? Fifteen minutes after taking the pill I think it would be pretty obvious to the person taking the pill whether or not caffeine was in the pill. (That's why millions of Americans drink a cup of coffee when they wake up in the morning.
Albert you seem to miss the point of double blind – the point isn't to blind a true physiological effect, even one that can be felt, but to blind to placebo.
Cam, The point of a double blind test is to hide from both the subject and the experimenter whether the subject is getting the treatment or the placebo (in an attempt to minimize placebo effects). I'm suggesting that in this case, attempting to hide that information from the subject is useless because 15 minutes after taking the pill, the subject knows exactly which pill he got.
Albert, sorry, it was a little silly to say you miss the point… sorry about that. I think both our perspectives make some sense. My argument would be this really is a double-blind trial: immediately after ingestion neither the experimenter nor the subject has any idea what pill they had, so any subsequent effect is a true effect of caffeine. I do get your point, but if the subject becomes aware of having taken caffeine precisely because the caffeine is having a physiological effect, isn't that actually what you're looking for (perhaps no… but it isn't far from it)?