In China, you get a cell phone number by buying a SIM card (a small plastic chip) that you put in your phone. Yesterday I bought one. I was shown a page of 12 possible numbers. At the top of the page it said 168 yuan ($25). But one of the numbers was cheaper: only 120 yuan ($18). Why the difference? I asked. The cheap number was “hard to remember,” I was told. I studied the 12 numbers; they looked equally hard to remember. So I got the cheap one.
Hard to remember was a euphemism, I learned later. Some digits (8, 6) are considered lucky, others (4, 7) unlucky. My number: 1170784.
This is related to self-experimentation. I suppose few scientists believe in lucky and unlucky phone numbers but many believe in “good” and “bad” ways of doing science. One example is a belief that self-experimentation is bad, another is a belief that Bayesian tools are “irrelevant to the business of science“; a third is the blue-ribbon panel that would only use data from double-blind experiments when deciding nutritional requirements. Scientists (and the rest of us) pay more than 48 yuan ($7) for such beliefs, which pervade science. Their effect is that scientists fail to use tools that would help them with their research; the rest of us suffer from the lack of progress that could have been made (e.g., discovery of better ways to treat depression). At the end of a paper about my self-experimentation I made this point:
Belief that something is bad makes it hard to learn what it is good for.