[R-meta] contour enhanced funnel plots when using function of sample size - R-SIG-meta-analysis

Wed, May 25, 2022 1:38 PM #

Hi Tori,

The funnel() function will work just as well on an rma.mv object, so yes, you can also create a contour-enhanced funnel plot for an rma.mv object. For example:

dat <- dat.konstantopoulos2011
res <- rma.mv(yi, vi, random = ~ 1 | district/school, data=dat)
funnel(res, level=c(90, 95, 99), shade=c("white", "gray55", "gray75"), refline=0, legend=TRUE)

Best,
Wolfgang

-----Original Message-----
From: R-sig-meta-analysis [mailto:r-sig-meta-analysis-bounces at r-project.org] On
Behalf Of Tori Pe?a
Sent: Wednesday, 25 May, 2022 21:41
To: r-sig-meta-analysis at r-project.org
Subject: Re: [R-meta] contour enhanced funnel plots when using function of sample
size

Hi folks,

I wanted to follow-up this question and ask if there is a way (or is it
recommended) to generate a color-enhanced funnel plot on a rma.mv object?
Thanks in advance!

On Tue, May 24, 2022 at 6:00 AM Guido Schwarzer <sc at imbi.uni-freiburg.de>
wrote:

Am 24.05.22 um 10:46 schrieb Brendan Hutchinson:

[...]
My question is specifically whether there is a means of generating a

contour enhanced funnel plot with pseudo confidence regions when using a
function of sample size on the y axis. According to the R documentation,
this isn't possible. I wanted to ask if anyone knew of any means to achieve
this or perhaps a workaround (unless this is some inherent issue to the use
of a function of sample size that I'm not seeing).

The sample size does not contain any information on the statistical
significance of a study result. For example, for a binary outcome:

 > meta::metabin(1, 100, 2, 100)$pval

[1] 0.5688381

 > meta::metabin(25, 100, 50, 100)$pval

[1] 0.0005287824

Both studies have a total sample size of 200, however, depending on the
number of events, the studies have very different p-values.

The same argument can be made for a continuous outcome, studies with the
same sample sizes and mean values can have very different standard
deviations leading to different p-values.

Best,

Guido

--
Dr. Guido Schwarzer
Institute of Medical Biometry and Statistics,
Faculty of Medicine and Medical Center - University of Freiburg

Postal address: Zinkmattenstr. 6a, D-79108 Freiburg

Mail: sc at imbi.uni-freiburg.de
Homepage: https://www.imbi.uni-freiburg.de/

ORCID iD: https://orcid.org/0000-0001-6214-9087
R-book: https://www.springer.com/gp/book/9783319214153