Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Friday February 16 2018, @02:27PM   Printer-friendly
from the character-assassination-for-dummies dept.

Argumentum ad hominem, a well-known fallacy that involves attacking the character or motive of the person making the argument rather than arguing their claims on their merits, is frequently encountered, and despite being fallacious, it is disturbingly effective. A new study in PLOS One (open, DOI: 10.1371/journal.pone.0192025) sheds some further light on just how effective the various types of ad hominem attacks are in the context of scientific claims. An article from Psypost reports on the findings:

Ad hominem arguments — attacking a person to disprove his or her claims — is considered a logical fallacy. But a new study published in PLOS One suggests that some ad hominem attacks can effectively erode people's trust in scientific claims.

The research found that attacking the motives of scientists undermines the belief in a scientific claim just as much as attacking the science itself.

[...] "One key finding is that if members of the general public are aware of a conflict of interest connected to a scientific finding, then this will seriously undermine their faith in that finding," Barnes told PsyPost. "What the study does is allow us to quantitatively compare the amount of attitude change based on knowledge of conflict of interest to the amount of attitude change based on knowledge of outright research fraud and misconduct (such as faking the data)."

"What we see is that knowledge of conflict of interest is just as powerful as knowledge of research fraud."

Further commentary on the study by Orac at Respectful Insolence.


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 4, Insightful) by meustrus on Friday February 16 2018, @04:17PM (7 children)

    by meustrus (4961) on Friday February 16 2018, @04:17PM (#638866)

    Ethics is a whole other can of worms. The problem here is that skepticism not based on the actual data is pre-empting actual analysis of the data. And sure, in the extreme case we have to take a really close look at the results of Nazi experiments trying to prove racial superiority, but then again we should be skeptical of the data anyway. It's quite possible to unethically produce interesting data, and whether its collection was ethical or unbiased should not affect our skeptical analysis of the data itself.

    Ad hominem is not just bad rhetoric when it's irrelevant. Ad hominem means attacking the credibility of the speaker, regardless of whether it's relevant to the words spoken. Ad hominem is always an argument we must ignore, because if we care more about the bias of the speaker than the words spoken, the actual data becomes irrelevant. Ad hominem leads to a world dominated by loyalty squabbles instead of objective analysis.

    --
    If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?
    Starting Score:    1  point
    Moderation   +2  
       Insightful=2, Total=2
    Extra 'Insightful' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   4  
  • (Score: 2) by HiThere on Friday February 16 2018, @07:28PM (2 children)

    by HiThere (866) Subscriber Badge on Friday February 16 2018, @07:28PM (#638962) Journal

    But if the speaker is untrustworthy, why should you trust their assertions?

    This is even true when they offer explicit data if you can't verify the provenance of the data. They could just be making it up. And, unfortunately, there have been many published scientific papers recently where the data *has* been either made up, or duplicated as additional confirmation.

    So the trustworthiness of the speaker is a valid consideration...at least as valid as the deductions that they make from the data they present, because if you can't trust the data, then the deductions, even if totally valid given the premises, are not trustworthy.

    Unfortunately, even this isn't the entire story, because of noise in the setup, etc. If failed experiments are suppressed, then you can't trust the result either. And this is a real problem, because there's no place where failed results are published, and failed experiments that should have succeeded were the desired hypothesis correct have scant impetus to publish. The relevant xkcd is https://xkcd.com/882/ [xkcd.com]
    OTOH, this has always been a problem, but when experiments were cheaper, rivals would publish experiments that refuted shaky assertions. These days, nobody's paying to do the refutations. So the errors are detected at a much later date...if ever.

    --
    Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
    • (Score: 2) by meustrus on Monday February 19 2018, @01:21AM (1 child)

      by meustrus (4961) on Monday February 19 2018, @01:21AM (#639899)

      Your approach essentially results in a loyalty test. Are you loyal to science? Are you loyal to your field? Are you loyal to a particular group of researchers? And are they loyal to you?

      Such a question should never be asked of scientists. Unfortunately these questions of loyalty have become the norm. And because scientists find themselves under assault as a group, they have become loyal to each other.

      Because scientists are loyal to each other, they find each other trustworthy. Because other scientists are trustworthy, their results must be valid. Therefore, judging results based on the trustworthiness of the speaker is the very reason why people no longer pay attention to the refutations.

      --
      If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?
      • (Score: 2) by HiThere on Monday February 19 2018, @05:34PM

        by HiThere (866) Subscriber Badge on Monday February 19 2018, @05:34PM (#640162) Journal

        You mean not having been caught lying is a loyalty test? I'm sorry, I can't take that as a serious response.

        Perhaps we mean different things by trustworthy. If someone is taking money to produce particular results, then I am less likely to believe them, and if they are taking money from a source that has a definite interest in particular results, then they are likely to be under pressure to produce those results. Etc. Calling doubting them "a loyalty test" is foolish, given the recent known examples of suppressed and dubious results.

        To doubt someone is not to assert that they are lying, it's to assert that there is insufficient evidence. It can be a good reason to look for more evidence, if you're interested enough, but it's not a good reason to decide. There are multiple reasons why one of the basis of science if independent replication, and one of them is to solve this problem. Unfortunately expense, transparency, and publication bias often mean that replication doesn't happen, which results in the promulgation of bad science. (Expense includes professional status, when you can't get your work published because "it's been done".)

        --
        Javascript is what you use to allow unknown third parties to run software you have no idea about on your computer.
  • (Score: 1) by khallow on Saturday February 17 2018, @01:50AM (3 children)

    by khallow (3766) Subscriber Badge on Saturday February 17 2018, @01:50AM (#639161) Journal

    Ad hominem is not just bad rhetoric when it's irrelevant. Ad hominem means attacking the credibility of the speaker, regardless of whether it's relevant to the words spoken.

    Which is worse? Attacking the credibility of a speaker with relevant concerns (which seems like to be the outcome of any criticism which involves considering those concerns) or ignoring those relevant concerns? There might be some middle ground here, but I don't have the nuance at present to see it.

    • (Score: 2) by meustrus on Monday February 19 2018, @01:27AM (2 children)

      by meustrus (4961) on Monday February 19 2018, @01:27AM (#639901)

      Those are not the only choices. The prerogative of science is that when the result is unclear, you find more data. You don’t decide whether or not to trust the opinion of the people that provided the incomplete data.

      --
      If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?
      • (Score: 1) by khallow on Monday February 19 2018, @01:44AM (1 child)

        by khallow (3766) Subscriber Badge on Monday February 19 2018, @01:44AM (#639907) Journal

        The prerogative of science is that when the result is unclear, you find more data.

        Such as data from parties that don't share the biases of the original group? That's the obvious way to get around conflicts of interest. Find groups with different interests to reproduce the results. But that takes years.

        This sort of independent data generation doesn't help the internet reader. They can't take a few minutes to generate their own data, independently confirming the research. Thus, consideration of heuristics like conflict of interest are a useful way to separate the wheat from the chaff.

        • (Score: 2) by meustrus on Monday February 19 2018, @02:27AM

          by meustrus (4961) on Monday February 19 2018, @02:27AM (#639926)

          I would never suggest that anybody trust anything on the internet at face value. But as a passive reader, yes, considering conflicts of interest is a pragmatic means of assessing bias. And while bias is not the same as invalid, most of the time the distinction is not important.

          When it is important, however, one cannot simply believe the author who is most likely to be correct. And if nobody seems to have the credibility to be correct, that doesn’t mean they aren’t.

          --
          If there isn't at least one reference or primary source, it's not +1 Informative. Maybe the underused +1 Interesting?