Anecdotal Evidence

[GUBU]

I have seen often here the claim; as put recently that anecdotal evidence is not evidence for anything more than the mundane. 

To my understanding at least, this seems to be used in an odd and at times seemingly forced use.  I am familiar; as one would peruse from a quick google search, the use of the term anecdotal evidence in a scientific sense.  A case such as: "Bob drank 8 glasses of water a day, and his cancer went away; therefore water cures cancer".  I don't dispute such uses of the terms or the reasoning.  Also, it seems that the use in such a case, it is not making a statement about the facts of the case (That Bob drank water, or was subsequently cancer free).  The issue here is that a general conclusion, is being made from what is normally a small sample size and insufficient reason.

So I would like for anyone interested: to clarify, what they mean by anecdotal evidence, particularly in regard to use in reference to Christianity.  Also the principle or justification of any claim in regards to evidence.

What you, and many people, don't get is that there is no point where anecdote becomes evidence. That is because anecdotes are based on personal memories, which are extremely fallible, due to having huge fallibility isseues, prejudice issues and unconscious bias issues (which apply equally to me as to thee). As a result they are not testable, verifiable or replicable, and therefore do not satisfy criteria needed ot qualify as evidence.

That being said anecdotes have a use in science, as they can point out interesting areas for further research, but for the research to be valid the initial anecdote has to be dropped, just as initial results from anything else that leads to a research study. The reason can be seen with card reading research in the 70s where initial high results weren't discarded for subsequent trials [initial high score predicters were invited back for further prediction tests], and people reverted to the mean in tests a lot more slowly than they should have. If the first results were dicarded for subsequent tests the results would have immediately been seen as random.