User Tag List

  1. #60581
    corona Hawthorne's Avatar
    Join Date
    Jan 2015
    MBTI
    INTP
    Enneagram
    5w4 so/sp
    Posts
    2,030

    Default

    Quote Originally Posted by uumlau View Post
    I have great skepticism of most statistical approaches to research, because instead of concluding (as they should) that "Hmm, these seem to be correlated. Research into phenomenology should look into this," they conclude, for example, "Cell phones are cooking your sperm!" (Mobile phones are 'cooking' men's sperm - Telegraph) Yes, a .2 Watt device is heating your sperm beyond acceptable levels, because statistics!
    To be fair, that has more to do with the way media people oversimplify and sensationalize stats than being an actual fault of the methods themselves.
    Likes yama liked this post

  2. #60582
    Sweet Summer Child yama's Avatar
    Join Date
    Dec 2014
    MBTI
    ESFJ
    Enneagram
    6w7 so/sx
    Socionics
    I'm Fe
    Posts
    7,508

    Default

    Quote Originally Posted by Sinclair View Post
    To be fair, that has more to do with the way media people oversimplify and sensationalize stats than being an actual fault of the methods themselves.
    Yup. And I'm taking social statistics (specifically qualitative), so we have a lot of emphasis on correctly interpreting data, not just doing route mathematics.
    MBTI: ESFJ
    Enneagram: 6w7 9w1 2w3 so/sx
    Temperament: Phlegmatic | Sanguine
    Astrology: Leo Sun | Aries Moon | Leo Rising
    Johari | Nohari

    not a type description

  3. #60583
    Happy Dancer uumlau's Avatar
    Join Date
    Feb 2010
    MBTI
    INTJ
    Enneagram
    953 sp/so
    Posts
    5,708

    Default

    Quote Originally Posted by Sinclair View Post
    To be fair, that has more to do with the way media people oversimplify and sensationalize stats than being an actual fault of the methods themselves.
    Quote Originally Posted by 21lux View Post
    Yup. And I'm taking social statistics (specifically qualitative), so we have a lot of emphasis on correctly interpreting data, not just doing route mathematics.
    Um, no. It is an actual fault of the methods themselves. There are innumerable "scientific" papers that are nothing more than a set of measurements with "p-values":
    https://www.sciencenews.org/article/odds-are-its-wrong
    The problem with P values: defining clinical vs. statistical significance | Public Health
    The problem with p values: how significant are they, really?
    On the scalability of statistical procedures: why the p-value bashers just don’t get it. | Simply Statistics
    http://www.perfendo.org/docs/BayesPr...onceptions.pdf

    As some of these links note, the problem is not with "p-values", but how many scientists naively use them. The meaning of particular statistical measures is very, very tricky - the same level of tricky as quantum mechanics or special relativity. One bad assumption ruins the whole thing. And bad assumptions are very human and very easy, because it will look like something is true and reasonably so. But the notion that there is something "true" is just as unlikely for these reasonable-looking conclusions as it is for these: http://tylervigen.com/spurious-correlations

    Random Thought Thread-chart-jpg

    chart.jpg
    An argument is two people sharing their ignorance.

    A discussion is two people sharing their understanding, even when they disagree.
    Likes SearchingforPeace liked this post

  4. #60584
    Sweet Summer Child yama's Avatar
    Join Date
    Dec 2014
    MBTI
    ESFJ
    Enneagram
    6w7 so/sx
    Socionics
    I'm Fe
    Posts
    7,508

    Default

    I should probably mention that I'm talking strictly about General Social Survey data (which still has its problems and should be taken with a grain of salt as to the accuracy of the respondents' answers). I can't speak for other types of data or sources because I don't use them.
    MBTI: ESFJ
    Enneagram: 6w7 9w1 2w3 so/sx
    Temperament: Phlegmatic | Sanguine
    Astrology: Leo Sun | Aries Moon | Leo Rising
    Johari | Nohari

    not a type description
    Likes uumlau liked this post

  5. #60585
    Happy Dancer uumlau's Avatar
    Join Date
    Feb 2010
    MBTI
    INTJ
    Enneagram
    953 sp/so
    Posts
    5,708

    Default

    Quote Originally Posted by 21lux View Post
    I should probably mention that I'm talking strictly about General Social Survey data (which still has its problems and should be taken with a grain of salt as to the accuracy of the respondents' answers). I can't speak for other types of data or sources because I don't use them.
    No worries. I'm just giving you a bit of a hard time.

    As long as you know the limitations of statistics, you're fine. It's just that in a publish-or-perish atmosphere in academia, there is a lot of incentive to publish half-assed papers that will pass peer review solely because they have the requisite p-values, and the reviewers don't much care about how the conclusions are interpreted as long as they aren't unreasonable.
    An argument is two people sharing their ignorance.

    A discussion is two people sharing their understanding, even when they disagree.

  6. #60586
    Sweet Summer Child yama's Avatar
    Join Date
    Dec 2014
    MBTI
    ESFJ
    Enneagram
    6w7 so/sx
    Socionics
    I'm Fe
    Posts
    7,508

    Default

    Quote Originally Posted by uumlau View Post
    No worries. I'm just giving you a bit of a hard time.

    As long as you know the limitations of statistics, you're fine. It's just that in a publish-or-perish atmosphere in academia, there is a lot of incentive to publish half-assed papers that will pass peer review solely because they have the requisite p-values, and the reviewers don't much care about how the conclusions are interpreted as long as they aren't unreasonable.
    Oh definitely. There's so much bullshit floating around that it's unbelievable.
    MBTI: ESFJ
    Enneagram: 6w7 9w1 2w3 so/sx
    Temperament: Phlegmatic | Sanguine
    Astrology: Leo Sun | Aries Moon | Leo Rising
    Johari | Nohari

    not a type description

  7. #60587
    corona Hawthorne's Avatar
    Join Date
    Jan 2015
    MBTI
    INTP
    Enneagram
    5w4 so/sp
    Posts
    2,030

    Default

    Quote Originally Posted by uumlau View Post
    Um, no. It is an actual fault of the methods themselves. There are innumerable "scientific" papers that are nothing more than a set of measurements with "p-values":
    https://www.sciencenews.org/article/odds-are-its-wrong
    The problem with P values: defining clinical vs. statistical significance | Public Health
    The problem with p values: how significant are they, really?
    On the scalability of statistical procedures: why the p-value bashers just don’t get it. | Simply Statistics
    http://www.perfendo.org/docs/BayesPr...onceptions.pdf

    As some of these links note, the problem is not with "p-values", but how many scientists naively use them. The meaning of particular statistical measures is very, very tricky - the same level of tricky as quantum mechanics or special relativity. One bad assumption ruins the whole thing. And bad assumptions are very human and very easy, because it will look like something is true and reasonably so. But the notion that there is something "true" is just as unlikely for these reasonable-looking conclusions as it is for these: http://tylervigen.com/spurious-correlations

    Random Thought Thread-chart-jpg

    chart.jpg
    I feel like the bolded is exactly what I was suggesting in my reply to your cellphone example. The tools, while not perfect, are only as effective as when properly utilized and rigorously interpreted but many people don't take the care to do so. Perhaps deliberately or simply out of ignorance. My "to be fair" was pointing out that much of the unreliability we see in articles is inaccurate interpretation or accurate interpretation of badly done statistics.

    This article you linked says exactly that and I noted that all the others focused on the way we analyze significance not the methods we use to measure it. Even the phys.org one didn't say the problem was with using p-values, it's the consistent use of the default 0.05 p-value when a more or less selective one would be better for the given experiment.

    tl;dr: I don't think we're in disagreement about the shoddy findings of many research studies. I'm just curious why you place the blame on statistics as a whole when "scientists who do statistics badly" seems the better target? Unless I misinterpreted your original post?

  8. #60588
    Sweet Summer Child yama's Avatar
    Join Date
    Dec 2014
    MBTI
    ESFJ
    Enneagram
    6w7 so/sx
    Socionics
    I'm Fe
    Posts
    7,508

    Default

    Got one of my ISTJs to answer my johari and nohari muahahaha fucking finally
    MBTI: ESFJ
    Enneagram: 6w7 9w1 2w3 so/sx
    Temperament: Phlegmatic | Sanguine
    Astrology: Leo Sun | Aries Moon | Leo Rising
    Johari | Nohari

    not a type description

  9. #60589
    Sweet Summer Child yama's Avatar
    Join Date
    Dec 2014
    MBTI
    ESFJ
    Enneagram
    6w7 so/sx
    Socionics
    I'm Fe
    Posts
    7,508

    Default

    It's almost time for the next avatar in my rotation but I'm just lovin' these little birdie kisses so much.
    MBTI: ESFJ
    Enneagram: 6w7 9w1 2w3 so/sx
    Temperament: Phlegmatic | Sanguine
    Astrology: Leo Sun | Aries Moon | Leo Rising
    Johari | Nohari

    not a type description

  10. #60590
    alchemist Legion's Avatar
    Join Date
    Sep 2014
    MBTI
    INFJ
    Posts
    1,862

    Default

    Got some marvelous shit cooking up for one of my uni classes... I reckon I could pull it off, too.

Similar Threads

  1. [SJ] SJ Random Thought Thread
    By Giggly in forum The SJ Guardhouse (ESFJ, ISFJ, ESTJ, ISTJ)
    Replies: 838
    Last Post: Yesterday, 08:05 PM
  2. [NT] Random thought thread (NT version)
    By yenom in forum The NT Rationale (ENTP, INTP, ENTJ, INTJ)
    Replies: 155
    Last Post: 10-22-2016, 04:00 PM

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
Single Sign On provided by vBSSO