[R-meta] overall effect size

[Lukas Dylewski]

Dear Wolfgang,

Thank you very much for the answers.
Really helped me your tips and codes.

Lukasz

wt., 30 kwi 2019 o 20:42 Viechtbauer, Wolfgang (SP) <
wolfgang.viechtbauer using maastrichtuniversity.nl> napisał(a):

> Dear Lukas,
>
> You could use the predict() function to compute the predicted average
> effect for a each level of 'status' when fixing 'logmass' to the mean of
> this moderator. The predict() function will also give you the SE. It will
> be something like:
>
> predict(res1, newmods = c())
>
> but without the output from res1, I cannot tell you what needs to go
> inside the c().
>
> Best,
> Wolfgang
>
> -----Original Message-----
> From: R-sig-meta-analysis [mailto:
> r-sig-meta-analysis-bounces using r-project.org] On Behalf Of Lukas Dylewski
> Sent: Saturday, 13 April, 2019 19:53
> To: r-sig-meta-analysis using r-project.org
> Subject: [R-meta] overall effect size
>
> Dear  Wolfgang ,
>
> I am working with a meta-analysis in R (metafor package) using a mixed
> model framework to examine experimental effects of rodent seed predation on
> plant recruitment as a function of various plant species characteristics. I
> decided to go with metafor given that this allowed to account for random
> factors (specifically the study source, each of which typically estimated
> effects for multiple species) while testing for the influence of multiple
> fixed factors in the same model.
>
> The one snag I’ve hit concerns deriving estimates of the mean effect size
> (the response) for my two classes of “status” (native or exotic) while
> accounting for the values of the continuous covariate “logmass.”  From
> playing with another meta-analysis package (OpenMEE, an interface that
> relies on the R package meta.analysis, which does not allow the addition of
> random factors).
>
> How I can estimate the mean effect size for "status" (alien and native)
> representative logmass values? Can I use the equation? But I’m unclear how
> I would calculate the associated SE for each estimate.  Could the solution
> possibly be as easy as a simple equation that you already have??
>
> The models look like:
>
> res1 <- rma.mv(yi, vi, mods = ~ logmass + I(logmass^2)+ factor(status) +
> logmass:factor(status), random=~1|pub,data=metaan, method="REML")
>
> Best
> Lukasz
> --
> Łukasz Dylewski, M. Sc.
> PhD Student
>
> Department of Zoology, Institute of Zoology
> Poznań University of Life Sciences
> Wojska Polskiego 71C
> 60-625 Poznań, Poland
>
> http://ecology1.wixsite.com/farmlandecology
>


-- 
Łukasz Dylewski, M. Sc.
PhD Student

Department of Zoology, Institute of Zoology
Poznań University of Life Sciences
Wojska Polskiego 71C
60-625 Poznań, Poland

http://ecology1.wixsite.com/farmlandecology

	[[alternative HTML version deleted]]