4 ms·

Scientists aren't idiots. Well, except for evolutionary biologists as described by Gould. :-) It is a hypothesis, to be tested. (I already wrote a comment argu

Scientists aren't idiots. Well, except for evolutionary biologists as described by Gould. :-) It is a hypothesis, to be tested. (I already wrote a comment arguing it is an unlikely hypothesis.) Edit: I might add, evol biologists routinely generate and test if e.g. genetic drift can be involved when they write papers.

fchollet 12y ago
> Scientists aren't idiots. You'd be amazed at how little most practicing scientists are versed in rational thinking, statistical rigorousness, and the scientific method. Maybe a lack of proper math and statistics education? I've since long lost count of the neuropsychology and biology papers I've come across that not only presented as fact results that did not really disprove the null hypothesis at all, but even advanced convoluted, ridiculous causal explanation for the (as likely inexistent) correlations they found.
- BugBrother 12y ago
  You find bad papers in every area. To criticize a paper's hypothesis which the authors explicitly show they didn't get good correlation for (and was surprised by), seems... well, a bit weird. Can you SHOW they are incompetent idiots? [Edit: At least that the paper is garbage, as you write?] You must have gotten that really quick, without reading their references? This is the end of the paper's discussion: "We proposed and tested a relatively new hypothesis to account for the presence of the Dark Triad as indicators of an evolved psychosocial strategy directed towards [...] We have provided a unique test, arguing and showing that those high on the Dark Triad may be characterized by cognitive biases that [..]" I agree that the paper is likely wrong -- but I see no proof the authors are a pack of incompetent f-ckwits, as you describe them as. As I wrote, they put forward a hypothesis and tried to test it.
  - fchollet 12y ago
    I'm not saying anyone is stupid. Rigorous statistical thinking is not particularly linked to intelligence. What I'm saying is: without rigorous statistical notions, good experimental science is impossible, and not nearly enough scientists have the required statistics and math background. Hence the banality of bad science among some experimental fields. Maybe you are a software developer (I don't know). In that case, you might have noticed that quite a few software developers around you are, well, incompetent at what they do. The same goes for managers and every other demanding profession. Scientists are no exception. In fact, to be a good scientist is actually much more difficult than to master a trade.
    - BugBrother 12y ago
      Despite three comments I see no motivation for your claim that the paper is obviously utter garbage. Edit: Removed the rest. Not needed. Bye. Edit 2: A general discussion of the quality of papers (which everyone with some time at U ought to know about) and a discussion of me personally is as typical a "changing the subject" strategy as it can be.

fchollet 12y ago

> Scientists aren't idiots. You'd be amazed at how little most practicing scientists are versed in rational thinking, statistical rigorousness, and the scientific method. Maybe a lack of proper math and statistics education? I've since long lost count of the neuropsychology and biology papers I've come across that not only presented as fact results that did not really disprove the null hypothesis at all, but even advanced convoluted, ridiculous causal explanation for the (as likely inexistent) correlations they found.

BugBrother 12y ago
You find bad papers in every area. To criticize a paper's hypothesis which the authors explicitly show they didn't get good correlation for (and was surprised by), seems... well, a bit weird. Can you SHOW they are incompetent idiots? [Edit: At least that the paper is garbage, as you write?] You must have gotten that really quick, without reading their references? This is the end of the paper's discussion: "We proposed and tested a relatively new hypothesis to account for the presence of the Dark Triad as indicators of an evolved psychosocial strategy directed towards [...] We have provided a unique test, arguing and showing that those high on the Dark Triad may be characterized by cognitive biases that [..]" I agree that the paper is likely wrong -- but I see no proof the authors are a pack of incompetent f-ckwits, as you describe them as. As I wrote, they put forward a hypothesis and tried to test it.
- fchollet 12y ago
  I'm not saying anyone is stupid. Rigorous statistical thinking is not particularly linked to intelligence. What I'm saying is: without rigorous statistical notions, good experimental science is impossible, and not nearly enough scientists have the required statistics and math background. Hence the banality of bad science among some experimental fields. Maybe you are a software developer (I don't know). In that case, you might have noticed that quite a few software developers around you are, well, incompetent at what they do. The same goes for managers and every other demanding profession. Scientists are no exception. In fact, to be a good scientist is actually much more difficult than to master a trade.
  - BugBrother 12y ago
    Despite three comments I see no motivation for your claim that the paper is obviously utter garbage. Edit: Removed the rest. Not needed. Bye. Edit 2: A general discussion of the quality of papers (which everyone with some time at U ought to know about) and a discussion of me personally is as typical a "changing the subject" strategy as it can be.