Blog Coordinator

X-Phi Grad Programs

« Philosophical Expertise? | Main | Even More Expertise! »



Feed You can follow this conversation by subscribing to the comment feed for this post.


Thanks for the paper. Can you please report effect sizes (preferably eta-squared or cohen's ds) and confidence intervals, at least on your main contrasts?

Thank you,

Eric Schwitzgebel

Hi Zach -- Sorry for the slow comment approval! I agree that effect sizes are important and CIs, especially for null effects. The article is already so loaded with numbers, we'll have to think carefully about whether to add a whole new set. For the percentage comparisons, effect sizes can be calculated by the reader with just the N's and the percentages; and for the regressions the e^betas are an effect size measure, and we put CIs around the crucial null results there. But I'll consult with Fiery. Maybe we should add a little more effect size detail.

The comments to this entry are closed.