[R-meta] Question on funnel plot interpretation

Michael Dewey ||@t@ @end|ng |rom dewey@myzen@co@uk
Wed Jul 5 14:09:37 CEST 2023


Dear Gabriel

Before you continue searching for small study effects I think a search 
for possible moderators of the effects would be more worthwhile. If you 
believe that the purpose of meta-analysis is to synthesise then the 
spread of values you have calls that goal into question. If you believ 
that the purpose of meta-analysis is to understand why effects differ 
then searching for possible moderators must be more helpful than looking 
for small study efects.

Michael


On 05/07/2023 12:31, Gabriel Cotlier wrote:
> 
> Ok, so in my case would it be better  to use for instance either inverse 
> n or alternatively variance,  is this correct?
> 
> Regarding the high variability in the range of my Fisher's z transformed 
> values you noticed before.
> This variability could be due to the fact that the values come 
> from different geographies, use of varied treatments and 
> different methods as well?
> Is there a quantitative measure or way that can help to give account of 
> such variability or explain it?
> 
> Thanks a lot again for your advice and guidance.
> Kind regards,
> Gabriel
> 
> On Wed, Jul 5, 2023 at 2:20 PM Yefeng Yang <yefeng.yang1 using unsw.edu.au 
> <mailto:yefeng.yang1 using unsw.edu.au>> wrote:
> 
>     There is a misunderstanding. Fishers Zr's sampling variance (or
>     error)  is not correlated with Zr. Correlation coefficent r is
>     correlated with is variance or error. So, if you transformed r to
>     Zr, as you said in early email, variance should be a better measure
>     of precision contrast to inverse n.
> 
>     Best,
>     Yefeng
>     ------------------------------------------------------------------------
>     *From:* Gabriel Cotlier <gabiklm01 using gmail.com
>     <mailto:gabiklm01 using gmail.com>>
>     *Sent:* Wednesday, 5 July 2023 21:10
>     *To:* Yefeng Yang <yefeng.yang1 using unsw.edu.au
>     <mailto:yefeng.yang1 using unsw.edu.au>>
>     *Cc:* R Special Interest Group for Meta-Analysis
>     <r-sig-meta-analysis using r-project.org
>     <mailto:r-sig-meta-analysis using r-project.org>>; Michael Dewey
>     <lists using dewey.myzen.co.uk <mailto:lists using dewey.myzen.co.uk>>
>     *Subject:* Re: [R-meta] Question on funnel plot interpretation
>     Thanks a lot Yefeng
>     Regards,
>     Gabriel
> 
> 
>     On Wed, Jul 5, 2023 at 2:03 PM Yefeng Yang <yefeng.yang1 using unsw.edu.au
>     <mailto:yefeng.yang1 using unsw.edu.au>> wrote:
> 
>         If you are doing an Egger's test using rma.mv <http://rma.mv>, 
>         use SE of Zr estimates as the predictor and look at the slope's
>         estimate and corresponding test. Also, add other important
>         predictors that might cause variations.
> 
>         Best,
>         Yefeng
>         ------------------------------------------------------------------------
>         *From:* Gabriel Cotlier <gabiklm01 using gmail.com
>         <mailto:gabiklm01 using gmail.com>>
>         *Sent:* Wednesday, 5 July 2023 20:35
>         *To:* Yefeng Yang <yefeng.yang1 using unsw.edu.au
>         <mailto:yefeng.yang1 using unsw.edu.au>>
>         *Cc:* R Special Interest Group for Meta-Analysis
>         <r-sig-meta-analysis using r-project.org
>         <mailto:r-sig-meta-analysis using r-project.org>>; Michael Dewey
>         <lists using dewey.myzen.co.uk <mailto:lists using dewey.myzen.co.uk>>
>         *Subject:* Re: [R-meta] Question on funnel plot interpretation
>         Dear Michael and Yefeng ,
> 
>         Thank you very much for the interesting observations and
>         orientation provided.
> 
>         - Regarding the strangeness of primary studies.
>         Yes indeed, it is something I noticed before that some studies
>         have a very low value or almost no correlation (0.001) while
>         others have a very high value almost maximum possible (close to 1).
>           All correlations included (n = 149)  are the result of
>         Fisher's z-to-r transformation, and original values of r also in
>         some cases present such extreme values.
>         The primary dataset of correlations were in some cases given in
>         the screened studies, but for the vast majority of
>         the correlations were calculated by myself as Pearsons's
>         product-moment correlation employing the values reported by the
>         different studies. These correlations (r) were later transformed
>         to z by means of Fisher's /r-to-z/ transform.
>         Studies are very diverse, coming from different geographies,
>         with varied types of treatments and applying different methods.
>         I do not know the reason for such variability of the range of
>         the correlations, but it would be interesting to have a test or
>         quantitative way to give account of such variation in the range.
> 
>         - Regarding the  the subjectivity of the interpretation of the
>         funnel plot and that I cannot use the function regtest()since is
>         I am using rma.mv <http://rma.mv>() object, I also run numerical
>         test for publication bias employing different predictors:
> 
>         1. sampling variance
>         2. inverse of sampling variance
>         3. standard error
> 
>         However since for each of the cases/predictors used (see below)
>         I got a full model result, I assume --may be wrongly--that the
>         value I should take as the numerical estimation of the
>         publication bised is the "intercept" of the model, is this correct?
>         Is there a given range that might serve as a proxy indicator of
>         potential publication bias?
> 
>         Code and model's results below.
> 
>         Thanks a lot.
>         Kind regards,
>         Gabriel
> 
> 
>         ## NUMERICAL TEST FOR PUBLICATION BIAS
> 
>         ## extending Egger's test to more complex models.
>         ## "regression test for funnel plot asymmetry".
> 
>         ## 1. using : the sampling variance
>         PubB<-rma.mv <http://rma.mv>(yi = yi,
>                       V = vi,
>                       mods =  vi,
>                       random = ~ 1 | Article / Sample_ID,
>                       data = dat,
>                       method = "REML")
> 
>         PubB
>         # Multivariate Meta-Analysis Model (k = 149; method: REML)
>         #
>         # Variance Components:
>         #
>         #            estim    sqrt  nlvls  fixed             factor
>         # sigma^2.1  0.8908  0.9438     72     no            Article
>         # sigma^2.2  2.1970  1.4822    149     no  Article/Sample_ID
>         #
>         # Test for Residual Heterogeneity:
>         # QE(df = 147) = 24617.3110, p-val < .0001
>         #
>         # Test of Moderators (coefficient 2):
>         # QM(df = 1) = 1.4381, p-val = 0.2304
>         #
>         # Model Results:
>         #
>         #          estimate      se     zval    pval ci.lb
>         <http://ci.lb>   ci.ub
>         # intrcpt    0.5620  0.2584   2.1743  0.0297    0.0554  1.0685  *
>         # mods      -6.7997  5.6701  -1.1992  0.2304  -17.9130  4.3135
>         #
>         # ---
>         #   Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
> 
> 
>         ## 2. using : the inverse of the sampling variance
>         PubB_1<-rma.mv <http://rma.mv>(yi = yi,
>                       V = vi,
>                       mods =  1/vi,
>                       random = ~ 1 | Article / Sample_ID,
>                       data = dat,
>                       method = "REML")
> 
>         PubB_1
>         # Multivariate Meta-Analysis Model (k = 149; method: REML)
>         #
>         # Variance Components:
>         #
>         #            estim    sqrt  nlvls  fixed             factor
>         # sigma^2.1  0.9176  0.9579     72     no            Article
>         # sigma^2.2  2.1797  1.4764    149     no  Article/Sample_ID
>         #
>         # Test for Residual Heterogeneity:
>         #   QE(df = 147) = 23656.2997, p-val < .0001
>         #
>         # Test of Moderators (coefficient 2):
>         # QM(df = 1) = 1.4469, p-val = 0.2290
>         #
>         # Model Results:
>         #
>         #         estimate      se    zval    pval ci.lb <http://ci.lb>
>            ci.ub
>         # intrcpt    0.0957  0.2595  0.3689  0.7122  -0.4129  0.6044
>         # mods       0.0040  0.0034  1.2029  0.2290  -0.0025  0.0106
>         #
>         # ---
>         #   Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
> 
>         ## 3. using : standard errors (square-root of the sampling
>         variances)
> 
>         PubB_2<-rma.mv <http://rma.mv>(yi = yi,
>                         V = vi,
>                         mods =  sqrt(vi),
>                         random = ~ 1 | Article / Sample_ID,
>                         data = dat,
>                         method = "REML")
> 
>         PubB_2
>         # Multivariate Meta-Analysis Model (k = 149; method: REML)
>         #
>         # Variance Components:
>         #
>         #            estim    sqrt  nlvls  fixed             factor
>         # sigma^2.1  0.8991  0.9482     72     no            Article
>         # sigma^2.2  2.1952  1.4816    149     no  Article/Sample_ID
>         #
>         # Test for Residual Heterogeneity:
>         # QE(df = 147) = 24489.6313, p-val < .0001
>         #
>         # Test of Moderators (coefficient 2):
>         # QM(df = 1) = 1.2528, p-val = 0.2630
>         #
>         # Model Results:
>         #
>         #           estimate      se     zval    pval ci.lb
>         <http://ci.lb>   ci.ub
>         # intrcpt    0.7801  0.4374   1.7834  0.0745  -0.0772  1.6375  .
>         # mods      -2.6473  2.3652  -1.1193  0.2630  -7.2829  1.9883
>         #
>         # ---
>         #   Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
> 
> 
>         On Wed, Jul 5, 2023 at 12:45 PM Yefeng Yang
>         <yefeng.yang1 using unsw.edu.au <mailto:yefeng.yang1 using unsw.edu.au>> wrote:
> 
>             Dear Gabriel
> 
>             Apart from Michael observation on data checking before
>             analyses (which is always a good practice), I add one.
>             The funnel plot is just a visual check of publication bias.
>             So the observations based on the funnel plots are inevitably
>             subjective - I mean you think that is "randomly scattering",
>             while others might think not. In contrast, Egger's test is a
>             more objective way to test the asymmetry of a funnel plot.
>             Regarding how to do it, you can try to find them in the
>             archives associated with this mailing list.
> 
>             Please be noted that whether none funnel plot and Egger's
>             test can indicate publication bias directly. But it is
>             common to assume the asymmetry of a funnel plot is caused by
>             publication bias (or more precisely. small study effects),
>             after accounting for heterogeneity.
> 
>             Best,
>             Yefeng
>             ------------------------------------------------------------------------
>             *From:* R-sig-meta-analysis
>             <r-sig-meta-analysis-bounces using r-project.org
>             <mailto:r-sig-meta-analysis-bounces using r-project.org>> on
>             behalf of Michael Dewey via R-sig-meta-analysis
>             <r-sig-meta-analysis using r-project.org
>             <mailto:r-sig-meta-analysis using r-project.org>>
>             *Sent:* Wednesday, 5 July 2023 19:31
>             *To:* R Special Interest Group for Meta-Analysis
>             <r-sig-meta-analysis using r-project.org
>             <mailto:r-sig-meta-analysis using r-project.org>>
>             *Cc:* Michael Dewey <lists using dewey.myzen.co.uk
>             <mailto:lists using dewey.myzen.co.uk>>; Gabriel Cotlier
>             <gabiklm01 using gmail.com <mailto:gabiklm01 using gmail.com>>
>             *Subject:* Re: [R-meta] Question on funnel plot interpretation
>             Dear Gabriel
> 
>             My interpretation looking at your plots is that you have a
>             very strange
>             set of primary studies. If the x-axis is really the z
>             transformation of
>             r then some of  the r are .999 and some 0.001 which seems
>             worthy of
>             investigation before looking further.
> 
>             Michael
> 
>             On 05/07/2023 09:14, Gabriel Cotlier via R-sig-meta-analysis
>             wrote:
>             > Hello all,
>             > 
>             > I have produced a funnel plot on the basis of an rma.mv <http://rma.mv>
>             > <http://rma.mv>() objectapplied to all the data set together
>             ( not
>             > subsetting  using moderators ) as follows:
>             > 
>             > image.png
>             > 
>             > 
>             > When looking at the figure I tried to think that maybe one of the 
>             > following two interpretations could be the correct one:
>             > 
>             > a.  There is a kind of random scattering of the effect sizes, therefore 
>             > no symmetry is found and thus publication bised is observed.
>             > b.  Given the randomness of the effect sizes distribution covering the 
>             > plot's space unevenly there is not a clear pattern that can indicate 
>             > publication bias is observed.
>             > 
>             > Is any of this interpretation the correct one?
>             > 
>             > Thanks a lot.
>             > Kind regards,
>             > Gabriel
>             > 
>             > 
>             > #### CODE. ######
>             > funnel_all <- rma.mv <http://rma.mv> <http://rma.mv <http://rma.mv>>(yi,
>             >                       vi,
>             >                       random = ~ 1 | Article / Sample_ID,
>             >                       data=dat)
>             > png(file = "funnel.png",
>             >      width = 250,
>             >      height = 200,
>             >      res = 600,
>             >      units = "mm")
>             > # par(mfrow = c(2, 1))
>             > 
>             > # full data
>             > f1 <- funnel(funnel_all,
>             >               yaxis = "seinv",
>             >               level = c(90, 95, 99),
>             >               ylim = c(1, 20),
>             >               shade = c("white", "gray55", "gray75"),
>             >               refline = 0,
>             >               legend = TRUE)
>             > mtext("A", side = 3, line = 0, adj = -0.13, cex = 2)
>             > 
>             > dev.off()
>             > 
>             > <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>> Virus-free.www.avg.com <http://Virus-free.www.avg.com> <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient <http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient>>
>             > 
>             > <#DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
>             > 
>             > _______________________________________________
>             > R-sig-meta-analysis mailing list @ R-sig-meta-analysis using r-project.org
>             <mailto:R-sig-meta-analysis using r-project.org>
>             > To manage your subscription to this mailing list, go to:
>             > https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis
>             <https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis>
> 
>             -- 
>             Michael
>             http://www.dewey.myzen.co.uk/home.html
>             <http://www.dewey.myzen.co.uk/home.html>
> 
>             _______________________________________________
>             R-sig-meta-analysis mailing list @
>             R-sig-meta-analysis using r-project.org
>             <mailto:R-sig-meta-analysis using r-project.org>
>             To manage your subscription to this mailing list, go to:
>             https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis
>             <https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis>
> 

-- 
Michael
http://www.dewey.myzen.co.uk/home.html



More information about the R-sig-meta-analysis mailing list