4 ms·

> The "scientific venue" has been corrupted That's a pretty strong accusation. Do you have any evidence? Cuddy herself, in the linked article, even asks: "Why

> The "scientific venue" has been corrupted That's a pretty strong accusation. Do you have any evidence? Cuddy herself, in the linked article, even asks: "Why not come to a conference or hold a seminar?" I don't think you're wrong about the lack of quality reviewing, but I also think that blogs and Facebook and articles in pop sci news venues are pretty shitty places to address the quality of editorial review. And your post demonstrates why: honest mistakes -- or even complete non-mistakes stemming from gaps in established best practices! -- turn into out-right accusations of full on corruption. The comments on NYT demonstrate the same problem from the other side: lots of gender politics when Gelman wasn't motivated by gender, implicitly or otherwise, but rather by the mass appeal of Cuddy's work. Needless to say, I also think you're massively over-stating the maliciousness of people like Cuddy and Fiske. > Science was being done sloppily, and Fiske was both a beneficiary of that and in a role where she could have prevented it. Simmons seems to agree with me that you're really over-reaching here: "The original study wasn’t particularly egregious," he said. "It was published in 2010 before anyone was thinking about this." There's a huge difference between doing science the way you were trained (but sub-optimally), and outright "corruption". There's also a HUGE difference between noticing a problem and constructively pushing for more rigor within a sub-community, and going after particular people within that community so harshly you destroy their careers. The former is great for science. The latter might possibly by great for improving rigor (I'm pessimistic and think it all comes down to incentives, which Gelman et al don't really do much to change). But equivocating bad methodological/statistical training/practice with unethical and even criminal ("corrupt") behavior? That sort of dishonest fire-and-brimstone talk is just as damaging as a lack of rigor.

conistonwater 9y ago
Just so you know, the basic statistical problem with this type of research was pointed out back in the 60's (http://andrewgelman.com/2016/05/06/needed-an-intellectual-history-of-research-criticism-in-psychology/ http://andrewgelman.com/2016/05/06/needed-an-intellectual-hi...), it's just that nobody listened.
- throwawayjava 9y ago
  My point is that: 1. failing to have read Meehl's critique doesn't amount to "corruption", 2. that calling people "corrupt" for having failed to read Meehl is not likely to be the most effective or the most civil path toward changing things, and 3. To the extent that discussion of scientific methodology on public forums causes or contributes to (2), that's a bad thing and perhaps suggests a fundamental cost of doing science in this manner.
  - conistonwater 9y ago
    I think you're using a different sense of the word "corrupt" from the one in the post you're replying to ("The "scientific venue" has been corrupted."). I know it's easy to lose nuance in internet discussions, but please read the posts above more carefully/charitably. Nothing Cuddy did was corrupt in any sense, only poor science.
    - throwawayjava 9y ago
      I re-read the comment, and disagree. The preceding sentence is: Fiske controls the venue and is (mis-)using the "scientific venue" to denounce her critics I'm having trouble finding a charitable way to read the claim that someone is misusing a scientific venue. And much more to the point, I doubt even more that the subject of those sentences (or the e.g. tenure/promotion committee of that subject) could be reasonably expected to apply such an abundance of charity.
      - conistonwater 9y ago
        Think "corrupted file", not "corrupt official". The key difference is that there is a primary purpose for their existence (a file stores data, a scientific process produces good and reliable findings), and something can be corrupted (no longer suitable for its purpose) without malice involved. To call an official corrupted, you need to allege malice, to say that a file is corrupted, you only need to show that it's not valid for its purpose. It's not a good choice of a word anyway, you could make exactly the same valid point without any of the ambiguity.
mattmcknight 9y ago
"That's a pretty strong accusation. Do you have any evidence?" Fiske is the journal editor. Fiske is allowing articles with weak stats to get published. Fiske is publishing her defense in that journal. Fiske is not publishing critical articles in that journal. That's corruption based on her position of authority. Cuddy just had sloppy stats and her results failed to replicate. Notice how you deceptively switch to Cuddy from Fiske in your response, because you have no point about Fiske.
- throwawayjava 9y ago
  > Fiske is publishing her defense in that journal APS Observer is a magazine, not a journal. Furthermore, Fiske was not the president of APS or an editor of the Magazine when that article was published. And even if she were, what's the problem? I see letters from the editor in most magazines I read on a very regular basis. Finally, the claim that Fiske was construing her opinion on conduct as a scientific fact is just... well. I'll refrain from calling you corrupt or deceptive because it was probably just a misunderstanding :-) > Notice how you deceptively switch to Cuddy from Fiske I actually have no idea what you're trying to say or what this refers to. Can you be more explicit?

conistonwater 9y ago

Just so you know, the basic statistical problem with this type of research was pointed out back in the 60's (http://andrewgelman.com/2016/05/06/needed-an-intellectual-history-of-research-criticism-in-psychology/ http://andrewgelman.com/2016/05/06/needed-an-intellectual-hi...), it's just that nobody listened.

throwawayjava 9y ago
My point is that: 1. failing to have read Meehl's critique doesn't amount to "corruption", 2. that calling people "corrupt" for having failed to read Meehl is not likely to be the most effective or the most civil path toward changing things, and 3. To the extent that discussion of scientific methodology on public forums causes or contributes to (2), that's a bad thing and perhaps suggests a fundamental cost of doing science in this manner.
- conistonwater 9y ago
  I think you're using a different sense of the word "corrupt" from the one in the post you're replying to ("The "scientific venue" has been corrupted."). I know it's easy to lose nuance in internet discussions, but please read the posts above more carefully/charitably. Nothing Cuddy did was corrupt in any sense, only poor science.
  - throwawayjava 9y ago
    I re-read the comment, and disagree. The preceding sentence is: Fiske controls the venue and is (mis-)using the "scientific venue" to denounce her critics I'm having trouble finding a charitable way to read the claim that someone is misusing a scientific venue. And much more to the point, I doubt even more that the subject of those sentences (or the e.g. tenure/promotion committee of that subject) could be reasonably expected to apply such an abundance of charity.
    - conistonwater 9y ago
      Think "corrupted file", not "corrupt official". The key difference is that there is a primary purpose for their existence (a file stores data, a scientific process produces good and reliable findings), and something can be corrupted (no longer suitable for its purpose) without malice involved. To call an official corrupted, you need to allege malice, to say that a file is corrupted, you only need to show that it's not valid for its purpose. It's not a good choice of a word anyway, you could make exactly the same valid point without any of the ambiguity.

mattmcknight 9y ago

"That's a pretty strong accusation. Do you have any evidence?" Fiske is the journal editor. Fiske is allowing articles with weak stats to get published. Fiske is publishing her defense in that journal. Fiske is not publishing critical articles in that journal. That's corruption based on her position of authority. Cuddy just had sloppy stats and her results failed to replicate. Notice how you deceptively switch to Cuddy from Fiske in your response, because you have no point about Fiske.

throwawayjava 9y ago
> Fiske is publishing her defense in that journal APS Observer is a magazine, not a journal. Furthermore, Fiske was not the president of APS or an editor of the Magazine when that article was published. And even if she were, what's the problem? I see letters from the editor in most magazines I read on a very regular basis. Finally, the claim that Fiske was construing her opinion on conduct as a scientific fact is just... well. I'll refrain from calling you corrupt or deceptive because it was probably just a misunderstanding :-) > Notice how you deceptively switch to Cuddy from Fiske I actually have no idea what you're trying to say or what this refers to. Can you be more explicit?