[R-meta] PET-PEESE Question - R-SIG-meta-analysis

Tue, Dec 5, 2023 11:53 PM #

Dear Tori,

vi refers to sampling variance in your code, right?

In that case, slight code adjustments seem necessary:
PET: mods = ~sqrt(vi) #standard error as a moderator
PEESE: PET: mods = ~vi #variance as a moderator


Further reading: 
https://bookdown.org/MathiasHarrer/Doing_Meta_Analysis_in_R/pub-bias.html#pet-peese


Best,
Lukasz

Lukasz Stasielowicz
Osnabr?ck University
Institute for Psychology
Research methods, psychological assessment, and evaluation
Lise-Meitner-Stra?e 3
49076 Osnabr?ck (Germany)
Twitter: https://twitter.com/l_stasielowicz
Tel.: +49 541 969-7735

On 06.12.2023 04:45, r-sig-meta-analysis-request at r-project.org wrote:
> Send R-sig-meta-analysis mailing list submissions to
> 	r-sig-meta-analysis at r-project.org
> 
> To subscribe or unsubscribe via the World Wide Web, visit
> 	https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis
> or, via email, send a message with subject or body 'help' to
> 	r-sig-meta-analysis-request at r-project.org
> 
> You can reach the person managing the list at
> 	r-sig-meta-analysis-owner at r-project.org
> 
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of R-sig-meta-analysis digest..."
> 
> 
> Today's Topics:
> 
>     1. PET-PEESE Question (=?UTF-8?Q?Tori_Pe=C3=B1a?=)
>     2. Re:  R-meta] add comma to numbers in forest plot from
>        metaprop objet (Zoukal Sofia)
>     3. Re:  R-meta] add comma to numbers in forest plot from
>        metaprop objet (Zoukal Sofia)
>     4. resources re. synthesis of cov/corr matrices (Sicong Liu)
>     5. Re: resources re. synthesis of cov/corr matrices (Mike Cheung)
> 
> ----------------------------------------------------------------------
> 
> Message: 1
> Date: Tue, 5 Dec 2023 17:59:31 -0500
> From: =?UTF-8?Q?Tori_Pe=C3=B1a?= <Tori.Pena at stonybrook.edu>
> To: R Special Interest Group for Meta-Analysis
> 	<r-sig-meta-analysis at r-project.org>
> Subject: [R-meta] PET-PEESE Question
> Message-ID:
> 	<CAEmuQj8Waiq4Eho6bGbuUuvh81Ls2c0kYKGYvBPjcjw7OQVTcg at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
> 
> Hello,
> 
> Thank you again for all your guidance.
> 
> I have been trying to conduct PET-PEESE analyses on several rma.mv objects
> by adding variance as a moderator.  The following code works well:
> 
> #pet-peese test
> m_1_pet <- rma.mv(yi = yi,
>                       V = vi,
>                       mods = ~ vi,
>                       slab = STUDY.ID,
>                       data = Relationship_TE_Occurrence,
>                       random = ~ 1 | STUDY.ID/EFFECT_SIZE_ID,
>                       test = "t",
>                       method = "REML")
> 
> summary(m_1_pet, digits=3)
> 
> However, when I try to conduct the PEESE part I receive this error message:
> "Model matrix not of full rank. Cannot fit model" for the following code:
> 
> Relationship$SE_Mod <- Relationship$vi^2
> 
> m_1_peese <- rma.mv(yi = yi,
>                       V = vi,
>                       mods = ~ SE_Mod,
>                       slab = STUDY.ID,
>                       data = Relationship_TE_Occurrence,
>                       random = ~ 1 | STUDY.ID/EFFECT_SIZE_ID,
>                       test = "t",
>                       method = "REML")
> 
> summary(m_1_peese, digits=3)
> 
> Does anyone have any insight into why the model would fit for the PET part
> but not the PEESE? I'm not sure how to fix this issue.
> 
> Thanks,
> Tori
>

Wolfgang Viechtbauer

Sun, Dec 10, 2023 11:36 AM #

There is another issue:

Relationship$SE_Mod <- Relationship$vi^2

but the dataset is called "Relationship_TE_Occurrence".

Best,
Wolfgang

-----Original Message-----
From: R-sig-meta-analysis <r-sig-meta-analysis-bounces at r-project.org> On Behalf
Of Lukasz Stasielowicz via R-sig-meta-analysis
Sent: Wednesday, December 6, 2023 08:53
To: r-sig-meta-analysis at r-project.org
Cc: Lukasz Stasielowicz <lukasz.stasielowicz at uni-osnabrueck.de>
Subject: Re: [R-meta] PET-PEESE Question

Dear Tori,

vi refers to sampling variance in your code, right?

In that case, slight code adjustments seem necessary:
PET: mods = ~sqrt(vi) #standard error as a moderator
PEESE: PET: mods = ~vi #variance as a moderator


Further reading:
https://bookdown.org/MathiasHarrer/Doing_Meta_Analysis_in_R/pub-bias.html#pet-
peese


Best,
Lukasz
--
Lukasz Stasielowicz
Osnabr?ck University
Institute for Psychology
Research methods, psychological assessment, and evaluation
Lise-Meitner-Stra?e 3
49076 Osnabr?ck (Germany)
Twitter: https://twitter.com/l_stasielowicz
Tel.: +49 541 969-7735

Hello,

Thank you again for all your guidance.

I have been trying to conduct PET-PEESE analyses on several rma.mv objects
by adding variance as a moderator.  The following code works well:

#pet-peese test
m_1_pet <- rma.mv(yi = yi,
                      V = vi,
                      mods = ~ vi,
                      slab = STUDY.ID,
                      data = Relationship_TE_Occurrence,
                      random = ~ 1 | STUDY.ID/EFFECT_SIZE_ID,
                      test = "t",
                      method = "REML")

summary(m_1_pet, digits=3)

However, when I try to conduct the PEESE part I receive this error message:
"Model matrix not of full rank. Cannot fit model" for the following code:

Relationship$SE_Mod <- Relationship$vi^2

m_1_peese <- rma.mv(yi = yi,
                      V = vi,
                      mods = ~ SE_Mod,
                      slab = STUDY.ID,
                      data = Relationship_TE_Occurrence,
                      random = ~ 1 | STUDY.ID/EFFECT_SIZE_ID,
                      test = "t",
                      method = "REML")

summary(m_1_peese, digits=3)

Does anyone have any insight into why the model would fit for the PET part
but not the PEESE? I'm not sure how to fix this issue.

Thanks,
Tori