Balancing between empiricism and reliance on anecdotes

So I learnt today that there's a limit to how long thread titles can be. Anyway...

Like most science fans, I advocate a fairly empirical approach to life. However, empiricism alone cannot provide sufficient factual basis for daily decision making. A certain amount of anecdotal evidence must be relied upon to fill the gaping chasms left behind by empirical observation.

For many years, I've tried to come up with some sort of algorithm, however incomplete, describing when I feel inclined to rely on anecdotes as opposed to empirical evidence.

Unfortunately, I've come up with next to nothing. Which is really quite disappointing, because I'm often rather articulate.

So, my question to you is: How do you leverage between "taking someone's word for it" and investigating with an objective(-ish) mindset?

Any factors you feel may be relevant, like involvement of money, person who's supplying the information, internet/irl, time of day, etc. would be nice.

I find myself trying to "take the human out of humanity" [1] sometimes, and it has led to all sorts of situations. Paradoxically, I just explained my point with an anecdote, however undetailed. I believe in a balance to things, but I try as much as possible to stay empirical in my observations.

1: my mother told me I do that, word for word.

Personally, I don't take things at face value. Human beings are inherently flawed, and the rarity at finding someone trustworthy enough to be of equal value as yourself is next to impossible. Another way of saying it is that I have trust issues.

For the most part, I cant explain how or why I can trust the people I put faith in. Experiences and time or relationships has nothing to do with it. Purely, it seems to be an inherent instict that I've developed over time. I just happen to "know" that someone will be useful. In otherwords, an eye for talent.

anecdote = one empirical data point

The questions you have to ask yourself about "relying on an anecdote" are the same as the questions you have to ask about how reliable your data set is. Plus, is this anecdote only memorable because it's something that happened once, because it was unusual and surprising and tagged by the teller's memory as important?

Relying on anecdotes is just relying on a really small data set. So, the question isn't an "either/or" scenario, it's "how many data points do you need to feel confident?"

I avoid anecdotes whenever I can.

Of course, sometimes that isn't possible. Not everything has been subjected to statistical analysis.

I will take someone's word for it if there is no evidence to the contrary. Expertise matters -- I will sooner believe a economist on TV talking about the economy than a random internet dude calling him an idiot. Between two opposing anecdotes, I will pick the one with whom I agree closest to (obviously).

Indeed, explaining one's behaviors and beliefs without anecdotes would be nigh impossible.



I think I see where you're coming from. Although, I would add that an anecdote "doubly-impurified" data set. When objectively "pure" data is received, one always has a layer of internal preconception that colors it. While adjusting for the presence of this layer is undeniably important, for many social situations, it is understandable to let it be. But every time you hear an anecdote, you're seeing data that's been processed twice, and warrants some "decoding".

For me the source of the anecdote is very important for this decoding process, basically, "how much do this person's analytical tendencies correspond to my own?"

If a coworker of mine (I work in a research lab) was to tell me that a technique he had tried sucked, that would be a more valuable data point to me than if a friend of mine mentioned that his/her ex was a jerk.

Objective perception - and pls forgive me the choice of words - is a bitch.

The human mind as it is now at this point of evolution is bound to do one of two things:
A) believe that something is true or
B) proof it.

Based on my knowledge and experience I can say that the human mind is set for putting things into known categories. This is both unfortunate and good for us. Good, because it makes us process things faster, unfortunate because our own perception is most definitively flawed.

Starting with the fact that our senses aren't perfect, it continues to how data is stored in short and long term conditions and then finally how much of this data is really accessible after it has been stored.

The only way that is known to mankind at this point to get objective data is the scientific method, but hey ... who has the time to put everything to test day in day out?

According to the pragmatic theory, truth is verified and confirmed by the results of putting one's concepts into practice.

My personal mindset has a tendency towards negative pragmatism, meaning that I'll never stop testing, since an idea or theory "could never be proved right, because tomorrow's experiment might succeed in proving wrong what you thought was right." (Richard Feynman)

But endless testing would overload our everyday activities, because it means that we are constantly testing everything that we ever learned. Starting with tasks as simple as brushing your teeth in the morning:
Do I really have to brush my teeth?
Do I brush my teeth sufficient enough?
Is this tooth paste the right choice?
Is this tooth brush the right choice?
...
...
...

So basically we have to rely to a certain point on ideas and theories we learned from other people and have faith in the scientific method. And yes, faith. To really know something, we have to go out there and try to prove ourselves that the scientific method delivers true data (or not).

Regarding anecdotes:
An anecdote is a short and amusing or interesting story about a real incident or person. It is almost always flawed with subjectivity, therefore useless as evidence of anything.

I treat them like eye witnesses in a trial as opposed to forensic evidence: In the absence of definitive proof to the contrary, I tend to accept the anecdote as to be true, until proven otherwise.