Guy Prochilo  

@GuyProchilo

PhD Candidate (exp 2020). I tweet statistics, open-science, & useful tools. Based . He/him

Vrijeme pridruživanja: siječanj 2013.

Tweetovi

Blokirali ste korisnika/cu @GuyProchilo

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

  1. Prikvačeni tweet

    I’m excited to announce my first first-author paper - published in ! “An Extended Commentary on Post-Publication Peer Review In Organizational Neuroscience”: Coauthored by , , , &

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    4. velj

    People are very happy to adopt treatments without proper RCTs. But when an RCT suggests a treatment they use does not work, they suddenly become very interested in methodological rigour.

    Poništi
  3. proslijedio/la je Tweet
    prije 12 sati

    I often hear journals say they can’t take the risk of implementing a new policy when bigger, more prestigious journals haven’t done it yet. Well, Nature now offers open peer review (publishing content of peer review). Ball’s in your court, smaller journals

    Poništi
  4. "Thank you for your interest in our paper. Please find attached the code for reproducing our findings" The code

    Poništi
  5. proslijedio/la je Tweet
    4. velj

    A brilliant paper by , , and Ian White on a principled way of doing sensitivity analyses. Here’s a very basic decision tree taken from the paper and I also post some interesting excerpts/examples in the tweets following

    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    1. velj

    Academia can drag you down because rejection is the norm, not the exception. You must celebrate every success, no matter how small. Set small goals and throw a mini-party for yourself when you meet them.

    Poništi
  7. Equivalence Testing for Psychological Research: A Tutorial (full-text):

    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet
    31. sij

    When you start a you are learning how to do research. You don't know how to do it so ask questions. Lots of questions. If you already knew how to do a PhD there wouldn't be any point in doing it. You are doing it to learn.

    Poništi
  9. : Cohen’s d is not robust. Algina et al (2005) provide a formula for a robust Cohen’s d based on 20% trimmed means. Easily compute d-robust using `bootES::bootES` in R. The authors suggest using a percentile bootstrap approach for CIs: install.packages("bootES")

    Poništi
  10. A confidence interval (CI) is defined by a "confidence procedure". For example, a 95% CI is an interval of values that are constructed in such a way to capture the population value on 95% of occasions in a long-run of hypothetical repetitions.

    Poništi
  11. proslijedio/la je Tweet
    29. sij

    Hooray! 5 years' work, some great reviewers and 'Best practice guidance for linear mixed-effects models in psychological science' is in press at Journal of Memory and Language. 50 free copies: TLDR: FOR GOODNESS SAKE REPORT MODELS & RESULTS IN FULL. 1/

    Prikaži ovu nit
    Poništi
  12. : Are you designing a between groups study & hypothesizing "no difference" between each group? Perform a power analysis to ensure you can detect a Cohen's d that is sufficiently bounded by 'smallness' (-0.20 > d > 0.20) using TOSTER: TOSTER::powerTOSTtwo()

    Prikaži ovu nit
    Poništi
  13. "... the fact that the data were collected using fMRI and can be rendered in dramatic fashion on an anatomically precise MRI may not help address the psychological or theoretical significance of the data." Full-text link:

    Poništi
  14. proslijedio/la je Tweet
    29. sij
    Odgovor korisnicima

    As far as I know, both Jasp and JAMOVI use afex in the back-end to do ANOVAs. So post-hoc tests for both should be based on the same model. Note that you can choose between: emmeans(..., model = "univariate") # default emmeans(..., model = "multivariate")

    Poništi
  15. : Repeated measures ANOVA is traditionally a horrible nightmare in R. The `afex` package changes this. Easily specify a rm-model that gives identical results to SPSS using `afex::aov_car`. Automatically applies the GG correction. install.packages("afex")

    Poništi
  16. Post-publication peer review, which in some cases challenges the conclusions of published work, will play a critical role in the development of sound theory & practice decisions that emerge from organizational neuroscience. More in:

    Poništi
  17. Avoid Cohen’s ‘Small’, ‘Medium’, and ‘Large’ for Power Analysis (pay-wall):

    Prikaži ovu nit
    Poništi
  18. A Practical Primer To Power Analysis for Simple Experimental Designs (full-text):

    Prikaži ovu nit
    Poništi
  19. When the haphazard decisions you made at the start of your PhD turn out just fine.

    Poništi
  20. An 'expected effect size' in a power analysis is a best guess based on the evidence available. If wrong, this can have huge implications for detecting an effect if it exists. Consider varying your effect size under different scenarios & examine how this affects power.

    Prikaži ovu nit
    Poništi
  21. Safeguard Power as a Protection Against Imprecise Power Estimates (full-text):

    Prikaži ovu nit
    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:

    ·