Documentation of prior for covariance of group-specific terms

[fweber144]

Summary:

There might be a typo in the documentation concerning the parameterization of the prior for the covariance of group-specific terms.

Description:

The documentation says:

In the case where 𝚺 is 1×1, this shape parameter is the cross-group standard deviation in the parameters and its square is the variance.

Shouldn't this read something like

In the case where 𝚺 is 1×1, 𝜏 is the cross-group standard deviation in the parameters and its square is the variance (so the Gamma prior with its shape and scale directly applies to the cross-group standard deviation in the parameters).

?

Reproducible Steps:

Not applicable.

RStanARM Version:

2.21.1

R Version:

Not applicable.

Operating System:

Not applicable.

[jgabry]

Just noticed nobody responded to this, sorry! I think you're correct that this is a typo and it should be referring to tau (whereas now it reads as if it's referring to the shape parameter of the gamma). @bgoodri Do you agree?

[bgoodri]

Yes

On Fri, May 7, 2021 at 4:00 PM Jonah Gabry @.***> wrote:

Just noticed nobody responded to this, sorry! I think you're correct that this is a typo and it should be referring to tau (whereas now it reads as if it's referring to the shape parameter of the gamma). @bgoodri https://github.com/bgoodri Do you agree?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/stan-dev/rstanarm/issues/502#issuecomment-834737913, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAZ2XKSRUHAJA2LIMCTKNFLTMRBEDANCNFSM4WCWD6ZA .

[jgabry]

Ok this is fixed on the master branch and on the website now. Thanks @fweber144!