3 ms·

Yes, learning everything in detail as an expert is impossible, we need experts, and because by definition we take their opinion at face value, we need some way

Yes, learning everything in detail as an expert is impossible, we need experts, and because by definition we take their opinion at face value, we need some way to judge their credibility. Of course we have those, i.e., "signing authorit[ies]" that validate that the opinion/["data"] of the expert is of high quality: they are generally called universities. (I find it amusing when programming jargon is applied to the real world outside of programming, but presented as new concept. Especially like where, where we have had institutions that accomplish the same thing as "signing authorities" for several thousand years).

scarejunba 8y ago
Ah no, you’ve misunderstood. Yours is a different argument from mine, though it is similar. You’re talking about the cost of being informed enough to understand stuff. I’m talking about trusting the raw data. The conclusions someone draws from data can be fully verified by another party (what you appear to have called opinions or “data” with quotes). The raw data itself cannot be, even by the most informed party. I can attempt replication but I’ll never be able to verify your data even if you opened your dataset to me. For instance, I may say that on being given drug A, patients’ self-reported pain reduced within two days, unless they’d been taking drug B for a related condition. Now, you can read my paper and see if my hypothesis is reasonable, if my experiment actually tests my hypothesis, if my sampling method is reasonable, if my results actually reject the hypothesis, etc. etc. The thing you can never check is whether I made up the numbers. Even if you were the foremost authority in the field, you can’t tell that from the paper or my data.
- speedplane 8y ago
  > Ah no, you’ve misunderstood. Yours is a different argument from mine, though it is similar. You’re talking about the cost of being informed enough to understand stuff. I’m talking about trusting the raw data. You're right, I was focusing on false analysis, not false raw data. But perhaps I overlooked it because it's rare to see original raw data cited in anonymous forum. Do you see this? Also, the very fact that it was cited anonymously, and also in an internet forum rather than a serious publication should raise enough alarm bells to distrust it, no? I agree with you that having false data spread through discourse is problematic. It's just not terribly common. What I see more often, is accurate raw data being contorted and misrepresented to promote some viewpoint. Mark Twain had a nice quote... facts are stubborn things, but statistics are more pliable.

scarejunba 8y ago

Ah no, you’ve misunderstood. Yours is a different argument from mine, though it is similar. You’re talking about the cost of being informed enough to understand stuff. I’m talking about trusting the raw data. The conclusions someone draws from data can be fully verified by another party (what you appear to have called opinions or “data” with quotes). The raw data itself cannot be, even by the most informed party. I can attempt replication but I’ll never be able to verify your data even if you opened your dataset to me. For instance, I may say that on being given drug A, patients’ self-reported pain reduced within two days, unless they’d been taking drug B for a related condition. Now, you can read my paper and see if my hypothesis is reasonable, if my experiment actually tests my hypothesis, if my sampling method is reasonable, if my results actually reject the hypothesis, etc. etc. The thing you can never check is whether I made up the numbers. Even if you were the foremost authority in the field, you can’t tell that from the paper or my data.

speedplane 8y ago
> Ah no, you’ve misunderstood. Yours is a different argument from mine, though it is similar. You’re talking about the cost of being informed enough to understand stuff. I’m talking about trusting the raw data. You're right, I was focusing on false analysis, not false raw data. But perhaps I overlooked it because it's rare to see original raw data cited in anonymous forum. Do you see this? Also, the very fact that it was cited anonymously, and also in an internet forum rather than a serious publication should raise enough alarm bells to distrust it, no? I agree with you that having false data spread through discourse is problematic. It's just not terribly common. What I see more often, is accurate raw data being contorted and misrepresented to promote some viewpoint. Mark Twain had a nice quote... facts are stubborn things, but statistics are more pliable.