3 ms·

That sounds terrible, I'm sorry for your friend. Systemically that's broken and represents the potential for massive abuse. That said, I assume you'd agree tha

That sounds terrible, I'm sorry for your friend. Systemically that's broken and represents the potential for massive abuse. That said, I assume you'd agree that anecdotes don't constitute data, and that we shouldn't develop strong beliefs on the backs of them. To assume the prevalence of something on the basis of loud anecdotes leads us to being easily manipulated -- in the absence of data I believe should keep our priors weak and seek better data.

btilly 13y ago
Actually I believe that our brains are wired to work from anecdote to theory, not from statistics to theory. In a small group this is a pretty good strategy. Where it breaks down in life is that the odds that we hear about an anecdote depend on the anecdote. This leads to all kinds of bias. I know the limits of my conscious analysis. I therefore have come to accept that there is value for me in filtering out "something that happened to someone I directly know" from "something I heard about". Since this happened to someone I directly know, it affects what I think. YMMV, and if you follow my rule, definitely should. But you'll at least understand why I might come to different conclusions than you do.
- hammerzeit 13y ago
  Epistemology is not my strong suit so please excuse me here, but how can you reasonably advocate for your positions in a worldview that specifically privileges personal experience? I tend to agree that we infer theory from anecdotes not data and even that that is a good thing. But don't we have a responsibility after developing a theory to evaluate it before making claims about it?
  - btilly 13y ago
    Epistemology is not my strong suit so please excuse me here, but how can you reasonably advocate for your positions in a worldview that specifically privileges personal experience? Epistemology is about how we know what we think we know. This is very separate from rhetoric, the art of trying to convince others with your words. I personally focus on epistemology, which includes clarifying what I believe, why I believe it, how strongly I believe it, and what different things I could reasonably believe in different circumstances. Knowing these things often leads me to conclude, "I believe X, you seem to believe Y, there is no way that I am going to change your beliefs." This seems to me to be a feature, not a flaw. I am unable to be friends with strongly religious types who are unable to muster the same feature. I tend to agree that we infer theory from anecdotes not data and even that that is a good thing. But don't we have a responsibility after developing a theory to evaluate it before making claims about it? In my world view, yes. If you see me failing to properly qualify what I claim, then please call me on it. That said, when I make a claim, I'm almost never going to give my full supportive reasoning for my world view. Life is too short.
vacri 13y ago
Data is only as good as the interpretation of it. You've just handwaved away a police homicide rate of an order of magnitude because the general homicide rate is also much worse. Other western democracies are bunched around a homicide rate of around 1.0/100k (Germany is 0.8), and the US is a clear outlier at 4.8. Having double the proportion of this significantly higher homicide rate is not something to claim as 'proportionate' and hence not a particular concern for the issue at hand. One wonders at what point you would consider it 'ridiculously disproportionate'? 10%? 20%?

btilly 13y ago

Actually I believe that our brains are wired to work from anecdote to theory, not from statistics to theory. In a small group this is a pretty good strategy. Where it breaks down in life is that the odds that we hear about an anecdote depend on the anecdote. This leads to all kinds of bias. I know the limits of my conscious analysis. I therefore have come to accept that there is value for me in filtering out "something that happened to someone I directly know" from "something I heard about". Since this happened to someone I directly know, it affects what I think. YMMV, and if you follow my rule, definitely should. But you'll at least understand why I might come to different conclusions than you do.

hammerzeit 13y ago
Epistemology is not my strong suit so please excuse me here, but how can you reasonably advocate for your positions in a worldview that specifically privileges personal experience? I tend to agree that we infer theory from anecdotes not data and even that that is a good thing. But don't we have a responsibility after developing a theory to evaluate it before making claims about it?
- btilly 13y ago
  Epistemology is not my strong suit so please excuse me here, but how can you reasonably advocate for your positions in a worldview that specifically privileges personal experience? Epistemology is about how we know what we think we know. This is very separate from rhetoric, the art of trying to convince others with your words. I personally focus on epistemology, which includes clarifying what I believe, why I believe it, how strongly I believe it, and what different things I could reasonably believe in different circumstances. Knowing these things often leads me to conclude, "I believe X, you seem to believe Y, there is no way that I am going to change your beliefs." This seems to me to be a feature, not a flaw. I am unable to be friends with strongly religious types who are unable to muster the same feature. I tend to agree that we infer theory from anecdotes not data and even that that is a good thing. But don't we have a responsibility after developing a theory to evaluate it before making claims about it? In my world view, yes. If you see me failing to properly qualify what I claim, then please call me on it. That said, when I make a claim, I'm almost never going to give my full supportive reasoning for my world view. Life is too short.

vacri 13y ago

Data is only as good as the interpretation of it. You've just handwaved away a police homicide rate of an order of magnitude because the general homicide rate is also much worse. Other western democracies are bunched around a homicide rate of around 1.0/100k (Germany is 0.8), and the US is a clear outlier at 4.8. Having double the proportion of this significantly higher homicide rate is not something to claim as 'proportionate' and hence not a particular concern for the issue at hand. One wonders at what point you would consider it 'ridiculously disproportionate'? 10%? 20%?