nick michalak

@nmmichalak

graduate student, methods maven , , and . interests: social perception, evolution, methods reform, , . he/him

Ann Arbor, MI
Vrijeme pridruživanja: listopad 2009.

Tweetovi

Blokirali ste korisnika/cu @nmmichalak

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @nmmichalak

  1. Prikvačeni tweet
    22. pro 2019.

    I wrote a new post: Logistic Regression in R Some key takeaways 1/

    Prikaži ovu nit
    Poništi
  2. 4. velj
    Poništi
  3. 20. sij
    Prikaži ovu nit
    Poništi
  4. 20. sij

    10 is more confident than "inconfident." 2 is ... 🤷‍♂️🤔

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet

    Another of my favorite traditions: Weird/best charts we made in 2019!!!

    Poništi
  6. 22. pro 2019.

    to the multiple logistic response surface. These takeaways are by no means new. They are simply useful things to know when interpreting the simple models I present in the post. I hope folks find this useful! 6/6

    Prikaži ovu nit
    Poništi
  7. 22. pro 2019.

    You can see this when you compare the multiple linear regression response surface 5/

    Prikaži ovu nit
    Poništi
  8. 22. pro 2019.

    (3) Partly because of the S-shape of the logistic function, the predicted values from multiple logistic regression depend on the values of all the predictors in the model, even when there is no true interaction. 4/

    Prikaži ovu nit
    Poništi
  9. 22. pro 2019.

    (2) However, linear regression often makes impossible predictions (probabilities below 0% or above 100%). 3/

    Prikaži ovu nit
    Poništi
  10. 22. pro 2019.

    (1) With regard to diagnostic accuracy (i.e., correctly classifying 1s and 0s), linear regression and logistic regression perform equally well in the simple cases I present in this post. 2/

    Prikaži ovu nit
    Poništi
  11. 14. pro 2019.

    If (especially controversial) claims in your preprint depend on your data* and analyses, you shouldn't post or promote your preprint without those. * assuming you can share data, summary statistics, covariance matrices, etc.

    Poništi
  12. proslijedio/la je Tweet
    12. pro 2019.

    I can't even tell you how many times I've had a first-year student from a rural or under-resourced school tell me they're terrified that they're already behind. Not getting the courses they need only compounds this disadvantage. We need to find a solution to this. Now.

    Prikaži ovu nit
    Poništi
  13. proslijedio/la je Tweet
    Odgovor korisniku/ci

    There is some experimental support for this: . It shows that social systems are stable only if it contains highly punitive individuals ready to punish invading cheaters. You need bad men to keep the bad men from the door.

    Poništi
  14. 12. pro 2019.
    Poništi
  15. proslijedio/la je Tweet
    10. pro 2019.
    Poništi
  16. proslijedio/la je Tweet
    10. pro 2019.

    "Some are reluctant to ask for anything they are not offered; others believe they should play hardball. In reality, you should aim to strike a balance between asking for the necessary resources to ensure your success while avoiding...alienating future colleagues."

    Poništi
  17. 10. pro 2019.
    Prikaži ovu nit
    Poništi
  18. 10. pro 2019.
    Prikaži ovu nit
    Poništi
  19. 10. pro 2019.

    Dug into this (Kevin is right, of course 🙂). I found this excellent tutorial with many examples, visuals, references. Hairy stuff 🐻. As always, plot your data folks📈

    Poništi
  20. 9. pro 2019.

    Degrees of Freedom at the Start of the Second 100 Years: A Pedagogical Treatise

    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·