Apply moderation to merged contrasts or to linear model only?

[Roman-Si]

Hi,

I have a question regarding when it is suggested to apply the Bayes moderation. In the vignette Comparing2Groups, the moderation is applied only to the linear model, and then the results of the ContrastsMissing are appended.
In the publication of prolfqua the moderation is performed after merging the different contrasts.
I go with the second method but wanted to confirm which method is recommended.

Thanks for your assistance!

[wolski]

Dear Roman,

Please stick with the paper's version —first merge and then moderate. I will update the Vignette to align it with the paper as soon as possible.
Thank you for pointing out the discrepancy.

Best wishes