My guess is that this only applies when using a categorical moderator, but with a continuous moderator you would want to keep the intercept as it is meaningful (i.e. tells you what expected effect size when your continuous moderator is 0)? If right, perhaps some quick clarification and maybe an example with a continuous moderator would help!
Hi folks,
Apologies if this isn't the place for this as it might be a more statsy question assuming it is not a mistake.
Could you clarify here and in the tutorial why the intercept is removed in the publication bias tutorial please (e.g. https://github.com/itchyshin/publication_bias/blob/main/R/Suppl_Info_Updated.Rmd#L1078)?
My guess is that this only applies when using a categorical moderator, but with a continuous moderator you would want to keep the intercept as it is meaningful (i.e. tells you what expected effect size when your continuous moderator is 0)? If right, perhaps some quick clarification and maybe an example with a continuous moderator would help!
Cheers Matt