@@dsquintana damn :o thanks! Is it frequent for you to highlight the values given by weightfunction correction in the abstract, for example? In the past I have used TaF, but never gave it too much focus through the paper.
This depends, but I’d be hesitant to treat the bias-corrected values as definitive, as they’re estimates. You *could* say something like, “the summary effect size estimate was statistically significant but a meta-analysis using selection models indicated evidence of publication bias, with a bias-corrected corrected effect estimate that was considerably smaller than the original estimate...”
Hi, thank you for your videos! They have been extremely useful. In a different video you say that the weight function attributes more weight to studies that are less likely to be published (p>0.05) and less weight to those more likely to be published (p
How come I will get an error saying 'Random effects model: non-finite finite-difference value [1]'? It also showed a warning note that at least one p-value contains no effect sizes. Thanks for your help!
not really a comment but I just want to thank you. Your explanations are so clear.
Thanks for the comment, I’m very happy to hear this 😀
Thank you very much for this tutorial on MA and adjusting for publication bias!
Thanks for the comment, glad this was useful
how to assess publication bias if there were only 5 studies included in a meta-analysis?
You can still use the same methods, although some are a bit more poorly suited if you don’t have many studies (e.g., weight selection models)
Thank you for all your amazing videos!
How can I calculate the standard error for the effect size r in R or Jasp?
Which is the formula for effect size standard error that JASP assumes we use? I have seen at least two different formulas around
Is that different (or that different) from the correction given by trim and fill? Thanks
Yes, a different approach than trim and fill, which should be avoided for assessing and adjusting for publication bias datacolada.org/30
@@dsquintana damn :o thanks! Is it frequent for you to highlight the values given by weightfunction correction in the abstract, for example? In the past I have used TaF, but never gave it too much focus through the paper.
This depends, but I’d be hesitant to treat the bias-corrected values as definitive, as they’re estimates. You *could* say something like, “the summary effect size estimate was statistically significant but a meta-analysis using selection models indicated evidence of publication bias, with a bias-corrected corrected effect estimate that was considerably smaller than the original estimate...”
Got it! Thanks for all the answers. :)
Glad to help!
Thank you so much for sharing Daniel.
Is there any instruction to calculate esse?
I just figured it out. Thanks Daniel!
@@jinggao6875 can you please tell us how you did it?
I would also like to know
Hi, thank you for your videos! They have been extremely useful. In a different video you say that the weight function attributes more weight to studies that are less likely to be published (p>0.05) and less weight to those more likely to be published (p
Don't suppose you have a link to the articles that you used here?
How come I will get an error saying 'Random effects model: non-finite finite-difference value [1]'? It also showed a warning note that at least one p-value contains no effect sizes. Thanks for your help!
I encountered the same problem here, not sure how come the weightfunct function cannot proceed after I specify the logRR and variances as arguments.
Hi, how do I calculate ESSE? Thank you!
too quiet