3 ms·

> Finally, I know that "we found RNA building blocks and B3" on a meteorite is less likely that contamination, so the story seems fishy and premature. Since yo

> Finally, I know that "we found RNA building blocks and B3" on a meteorite is less likely that contamination, so the story seems fishy and premature. Since you rightly encourage everyone to question claims, I question this claim. How do you know this, specifically? In other words, we have a situation where a team has done a bunch of experiments, data collection, and analysis, and then published a paper claiming "we found RNA building blocks and B3", right? And now your claim is "no, you're wrong about that," but without any experiments, data collection, and analysis. So if you're going to make a claim, provide the details, data, and evidence that substantiates your claim.

coldtea 4y ago
You can call it a prior (like in Bayesian probability) on the probability of such an announcement being right. Same, to give an obvious extreme example, as I'd be sceptical about an annoucencement that "We found the perpetual machine, there was a major flaw in thermodynamics laws, here are our numbers and results", even if they're world reknown experts in physics saying it, and I don't have access to their data, or to the machine to try it myself - hell, even if I can't understand their formulas. One would still have a prior in this case, based on a general understanding of the the world, and the possibility range where this falls. You might have a more informed prior because you took more concrete information into account than mere experience of other such announcements being repeatedly proven wrong, the probability of RNA being found on a meteor sample, the probability of the team just being lax with their annoucement to get publicity and grants, and so on. Or, I guess, you think that we should never do that, and we should either do the research ourselves, or wait until someone else does, and at any point, whatever they say as their conclusion, them being experts and having done the work and everything, should be the final word. Which probably means that as you see it, it makes no sense, to have a belief on anything or its refutation as likely or unlikely, unless one has either replicated the study and checked the conclusions, or found specific actionable faults.
- craftsman 4y ago
  It's fine and quite legitimate to have a prior, as your example of a perpetual motion machine illustrates. Along with the prior it's also important to understand how typical cognitive biases can affect those priors. For example, you cite the "mere experience of other such announcements being repeatedly proven wrong" and "the probability of the team just being lax with their annoucement to get publicity and grants." These might be examples of availability bias. Clearly these things happen. The question is whether they actually happen with a frequency that would justify the statement: > Finally, I know that "we found RNA building blocks and B3" on a meteorite is less likely that contamination, so the story seems fishy and premature. Regardless of each of our own priors and biases, respectfully, you don't know this. You don't have evidence that shifts the probability in this direction. Or rather, you haven't presented any. You haven't pointed out a specific flaw in the experimental design, instrument design and operation, data collection and analysis. What you have done is state your strong prior suspicion that something is wrong. That's not knowledge, and it's nothing that anyone else could base their assessment upon. The statement is simply a statement of your prior, nothing else. It's not a material statement about the actual paper, its data, analysis, or its conclusion.

coldtea 4y ago

You can call it a prior (like in Bayesian probability) on the probability of such an announcement being right. Same, to give an obvious extreme example, as I'd be sceptical about an annoucencement that "We found the perpetual machine, there was a major flaw in thermodynamics laws, here are our numbers and results", even if they're world reknown experts in physics saying it, and I don't have access to their data, or to the machine to try it myself - hell, even if I can't understand their formulas. One would still have a prior in this case, based on a general understanding of the the world, and the possibility range where this falls. You might have a more informed prior because you took more concrete information into account than mere experience of other such announcements being repeatedly proven wrong, the probability of RNA being found on a meteor sample, the probability of the team just being lax with their annoucement to get publicity and grants, and so on. Or, I guess, you think that we should never do that, and we should either do the research ourselves, or wait until someone else does, and at any point, whatever they say as their conclusion, them being experts and having done the work and everything, should be the final word. Which probably means that as you see it, it makes no sense, to have a belief on anything or its refutation as likely or unlikely, unless one has either replicated the study and checked the conclusions, or found specific actionable faults.

craftsman 4y ago
It's fine and quite legitimate to have a prior, as your example of a perpetual motion machine illustrates. Along with the prior it's also important to understand how typical cognitive biases can affect those priors. For example, you cite the "mere experience of other such announcements being repeatedly proven wrong" and "the probability of the team just being lax with their annoucement to get publicity and grants." These might be examples of availability bias. Clearly these things happen. The question is whether they actually happen with a frequency that would justify the statement: > Finally, I know that "we found RNA building blocks and B3" on a meteorite is less likely that contamination, so the story seems fishy and premature. Regardless of each of our own priors and biases, respectfully, you don't know this. You don't have evidence that shifts the probability in this direction. Or rather, you haven't presented any. You haven't pointed out a specific flaw in the experimental design, instrument design and operation, data collection and analysis. What you have done is state your strong prior suspicion that something is wrong. That's not knowledge, and it's nothing that anyone else could base their assessment upon. The statement is simply a statement of your prior, nothing else. It's not a material statement about the actual paper, its data, analysis, or its conclusion.