4 ms·

New rule suggestion. You must argue at least one flaw in the study or article before pointing out the source may be biased.

by s_m_t 10y ago

simplemath 10y ago
And that argument should be supported by something other than pure conjecture
codingdave 10y ago
I really don't think HN is shy about pointing out flaws in any study shared here.
dstroot 10y ago
Ha! I had had that exact same reaction. Reading the actual document it comes across as factual and not biased and all sources are referenced.
foobarian 10y ago
I like to think of it as the US justice system. I no more expect the Cato institute to argue for outlawing marijuana use than I would expect a prosecutor to drop the case against an axe murderer with 10 witnesses and DNA evidence.
necessity 10y ago
One would think HN commentators would know Graham's hierarchy of disagreement...
c3RlcGhlbnI_ 10y ago
Why? Many of the flaws that jump out at me are not so significant that they would be worthy of concern on their own and I appreciated other people clarifying the group's affiliations(I probably wouldn't have been curious enough to go back and read a bit more otherwise). Though honestly if this is a scientific study it is very strange that they omitted their methodology entirely(how did they select the data, how did they compare the statistics before and after, how did they account for confounding factors, etc). That is pretty fundamental to this kind of analysis and in its absence I would be hesitant to put much faith in even their weak conclusions. The graphs themselves are definitely interesting though and I would not at all be surprised if they were on the right track.
- AstralStorm 10y ago
  If they jump out at you, it should be trivial to argue them, right? Thus the rule is no real obstruction to dialogue. Lack of methodology is a valid argument, as opposed to just pointing out the funding. Why not contact them about it? Or perhaps they have used some base sources, which could be made available to make their argument even stronger when exposed.

New rule suggestion. You must argue at least one flaw in the study or article before pointing out the source may be biased.

by s_m_t 10y ago

New rule suggestion. You must argue at least one flaw in the study or article before pointing out the source may be biased.

simplemath 10y ago

And that argument should be supported by something other than pure conjecture

codingdave 10y ago

I really don't think HN is shy about pointing out flaws in any study shared here.

dstroot 10y ago

Ha! I had had that exact same reaction. Reading the actual document it comes across as factual and not biased and all sources are referenced.

foobarian 10y ago

I like to think of it as the US justice system. I no more expect the Cato institute to argue for outlawing marijuana use than I would expect a prosecutor to drop the case against an axe murderer with 10 witnesses and DNA evidence.

necessity 10y ago

One would think HN commentators would know Graham's hierarchy of disagreement...

c3RlcGhlbnI_ 10y ago

Why? Many of the flaws that jump out at me are not so significant that they would be worthy of concern on their own and I appreciated other people clarifying the group's affiliations(I probably wouldn't have been curious enough to go back and read a bit more otherwise). Though honestly if this is a scientific study it is very strange that they omitted their methodology entirely(how did they select the data, how did they compare the statistics before and after, how did they account for confounding factors, etc). That is pretty fundamental to this kind of analysis and in its absence I would be hesitant to put much faith in even their weak conclusions. The graphs themselves are definitely interesting though and I would not at all be surprised if they were on the right track.

AstralStorm 10y ago
If they jump out at you, it should be trivial to argue them, right? Thus the rule is no real obstruction to dialogue. Lack of methodology is a valid argument, as opposed to just pointing out the funding. Why not contact them about it? Or perhaps they have used some base sources, which could be made available to make their argument even stronger when exposed.