[R-meta] Calculation of p values in selmodel

[Viechtbauer, Wolfgang (NP)]

See: https://www.metafor-project.org/doku.php/installation#development_version

> -----Original Message-----
> From: R-sig-meta-analysis <r-sig-meta-analysis-bounces using r-project.org> On Behalf
> Of Will Hopkins via R-sig-meta-analysis
> Sent: Wednesday, March 20, 2024 23:01
> To: 'R Special Interest Group for Meta-Analysis' <r-sig-meta-analysis using r-
> project.org>
> Cc: Will Hopkins <willthekiwi using gmail.com>
> Subject: Re: [R-meta] Calculation of p values in selmodel
>
> Wolfgang, I figured that I might be able to implement your selmodel(...,
> pval=dat$pValue) by reverting to for-loop processing. I solved the previous
> problems I had with this approach by stripping out the NULLs from the list
> objects, so the simulations are running nicely.
>
> However, it appears that the latest version of metafor, which I have just
> installed, does not have the pval functionality yet, because I get this warning
> with every invocation of selmodel:
> Extra argument ('pval') disregarded.
>
> I am still a newbie with R, so I may have done something wrong with the update
> process. I got this warning, and nine others like it, but I presume this doesn't
> apply to metafor:
> Warning: package 'boot' in library 'C:/Program Files/R/R-4.3.2/library' will not
> be updated.
>
> I then got this, again presumably irrelevant:
> Warning: cannot remove prior installation of package ‘lattice’
> Warning: restored ‘lattice’
>
> The only warning for loading metafor was this:
> package ‘Matrix’ was built under R version 4.3.3
>
> The other possibility is that my for-loop processing has some kind of mistake,
> but it seems to be OK. Each loop produces a meta-analysis object resultmeta from
> invocation of ram.uni with a dataset meta_data (the i-th subset of my full
> dataset). I then apply selmodel to resultmeta within a function, where the x is
> resultmeta:
> result <- metafor::selmodel(x,  type="step", steps=(0.025),
> pval=meta_data$pValue)
>
> I give up. Help, please!
>
> Will