The Tendency to Find Something

I wanted to say a few words about how scientists report their work.  This week's podcast of EconTalk caused me to think about this issue a bit.

In this episode the host pointed out that when scientists rarely write about what they actually do.  They don't tell you about all of the things that they did that turned out not to back up their findings.  I think this is the case.  I have either heard or read that generally when a scientist comes up with a surprising result, that the surprising result turns out to be wrong.  Scientists are human after all, end hence prone to error.  If scientists are trying to do something extremely complicated, then there is even more opportunity for them to have made some sort of mistake.

However, this can lead to problems.  If scientists omit the times that they did the experiment and got the wrong result, then there is a potential for them to tweak things until the results conform to their prior expectations.