Listening to Freakonomics Radio on lottery-like savings accounts reminded me of a big reason I self-experiment: it resembles buying a lottery ticket. Whenever you collect data I believe there is a power-law distribution of benefit: large chance of little benefit, small chance of large benefit. (Your sophistication and other things affect the slope.) Almost all data confirms what you already knew — small benefit. A very small fraction of data gives you a new idea — large benefit. Because self-experimentation is about oneself, new ideas can have tangible benefits, just as winning the lottery provides tangible benefit (money)
Basically I hope for outliers — a sudden jump up or down in something I’m tracking, such as arithmetic speed or sleep duration. This may give me a new idea about what controls that measure. Self-experiments are also valuable because something I’m not deliberately measuring may change. When I started watching faces on TV in the morning to see if it would affect my sleep, my mood, which I wasn’t deliberately measuring, suddenly improved.
It really does feel like playing the lottery for free. To not make a measurement I could easily make feels like walking by a perfectly good lottery ticket lying in the street. Loss aversion sets in.
Hi Seth,
I love your approach of self-experimentation and I’m a big advocate for it.
I don’t know if the relationship between male testosterone levels and abstinence has been brought to your attention. People who abstain report improved brain function, strength, mood, overall energy and focus. A Chinese study found that men’s T peaked after 7 days post-ejaculation [the study seems to have stopped soon after, there is no data it seems regarding abstaining for longer periods].
Similar to your SLD forums, there is this massive forum thread in which hundreds of guys testify to these effects:
https://www.sosuave.net/forum/showthread.php?s=060780d57fe8bcad7e90e782d797c32c&t=20053
It would be great if you could give this topic attention. Many people dismiss it since it doesn’t go well with popular belief, but in my opinion the experimental evidence of all of these forum posters is too much to ignore.
Data collection always seems like such a lot of work to me. You explained very well why people do it. It reminds me of my own twisted relationship with Google Reader. Except theres a large chance I read mostly mundane stuff. There’s always those times where I learn something really interesting or useful. Not too often but enough to keep me, what, addicted? I’ll never admit to that.
I expect that the random timing of the big benefits strengthens the motivation.
Mark, thanks, I didn’t know about that.
“the random timing of the big benefits strengthens the motivation.” Good point. Learning psychologists such as me know very well that partial reinforcement (e.g., rewarding only 25% of bar presses) produces great resistance to extinction than 100% reinforcement.
Seth, what do you see as the path from your (or anyone’s) initial self-experimentation all the way to acceptance by the larger scientific community? In previous comments, you seem to have down-played the importance of replicating your results — or, at least, of scaling-up studies to use larger sample sizes. So… where is one to go after discovering something new in oneself?
Alex, the path to wider acceptance is to find out if it is true in other people. I believe that is best done by starting small, the smaller the better. It is a poor idea to do big studies before you have done small studies. That doesn’t mean big studies are bad or unimportant; it means small studies should be done first. Maybe my preference for starting small is why you interpreted earlier comments by me as “downplaying the importance of replicating [my] results” which is not true at all. I think it is very important to learn to what extent the results can be repeated.