“Correlation does not equal causality” is an often-heard phrase when someone says “I did X and Y happened.” It occurred to me a while ago that we should stop saying it that way.
If I do a thing one time and get a certain result, that’s just a coincidence. One data point is not enough to even begin to establish a correlation.
This is on the same level as anecdotal evidence versus statistical evidence. One time is an anecdote or a coincidence; many times under different circumstances involving different people is statistics and correlation.
Eliminate the variables that can logically explain or account for the correlation and you have a reasonable theory about causes. Then, if it’s testable, you have a hypothesis.
Far too many people stop at the level of anecdotal or coincidental and think they know something. This can be seen in many of the books written by so-called self-made men: they don’t have more than their own personal experiences to go on, so that’s all they know.
It is possible to do research and compile a collection of results (called a survey) that compares these anecdotal reports to see if their knowledge applies beyond their coincidental situation. Actually converting such a survey into a testable hypothesis is very difficult in many areas of life, which is why we get computer modeling presented as if it were science.
Computers can be very useful and they can do amazing things, but they’re limited in ways that actual real-world testing and experience are not.
It’s human nature to shy away from what’s difficult, so I don’t blame people for doing that. I do think we need to recognize that facts and reality don’t care whether we find them difficult or inconvenient. They’re still going to affect us, no matter how we feel about them or even if we ignore or deny them.
My point is that your experience is valid, as are your feelings about that experience — but it tells us nothing about causes, except in a very limited way.