[R-meta] question on scatter plot of estimates (Fisher's Z) against the standard error

Michael Dewey ||@t@ @end|ng |rom dewey@myzen@co@uk
Thu Jul 20 15:51:15 CEST 2023


Dear Gabriel

Comments in-line

On 20/07/2023 05:55, Gabriel Cotlier wrote:
> Dear Michael,
> 
> I think you are completely right, in the fact, the plot I am producing 
> is indeed valid for the purpose for which I want to use it, meaning it 
> is representative of the relationship I want to show. Therefore, I 
> assume that the plot I am getting, is supposed to be sufficient.
> 
> However, I receive from the function metafore:: funnel (model), for a 
> model without modierators, a very nice representation of the scarring of 
> the observed outcomes or the estimates (x axis), as a function of the SE 
> (e.i., square root of the sampling variance, SE assumef to have a pseudo 
> confidence interval region drawn around each of its values). While, when 
> I plot by myself
> x = observed outcomes
> y = square root of the sampling variance,
> 
> Then the plot shows that:
> a.  the scattering of the points appears upside down with respect to the 
> output of the function metafore:: funnel (model),

I have already answered that one in a previous post. It is just the 
convention

> b. the scale of the y axis, instead of having a defined top at zero and 
> from there values are represented downwards, the scale is different.
> 

Without your code it is hard to tell but I suspect you are not plotting 
what you think you are. Are you plotting the inverse of the se?

Michael

> 
> Anyways, I started thinking that in any case, such a difference in the 
> plot I am doing by myself is not necessarily wrong, but is just a 
> different way of representing the data. Just the scattering of the 
> points in one  case looks like the upside down scattering of the other. 
> And I assume this is because maybe the function metafore::funnel() 
> applies some operation on the square root of the mean (y axis) that I 
> presume is the calculation of the aforementioned pseudo confidence 
> interval for each value, but I am not sure.
> 
> Thanks a lot for your response.
> Kind regards,
> Gabriel
> 
> On Wed, Jul 19, 2023 at 7:20 PM Michael Dewey <lists using dewey.myzen.co.uk 
> <mailto:lists using dewey.myzen.co.uk>> wrote:
> 
>     I am sorry Gabriel but I do not understand why the plot you say you
>     produced fails to do what you say you want.
> 
>     Michael
> 
>     On 19/07/2023 10:59, Gabriel Cotlier wrote:
>      > Hello Michael,
>      > Thank you very much for your response.
>      > I just would like to show that the of data set I have has high
>      > uncertainty given that no possible pattern is observable or
>     detectable
>      > and no order is possible to visulize in the scattering,
>      > I thought that a plot with x axis = fisher's z observed
>      > outcomes (estimates) and y axis = standard error or any
>     other measure of
>      > uncertainty could at least visually demostrate that assumption.
>      > If such a lack of pattern or high uncertainty in the data set can
>     also
>      > be demonstrated numerically, even better.
>      > Kind regards,
>      > Gabriel
>      >
>      > On Wed, Jul 19, 2023 at 12:29 PM Michael Dewey
>     <lists using dewey.myzen.co.uk <mailto:lists using dewey.myzen.co.uk>
>      > <mailto:lists using dewey.myzen.co.uk
>     <mailto:lists using dewey.myzen.co.uk>>> wrote:
>      >
>      >     Dear Gabriel
>      >
>      >     I am not realy sure what you are trying to do but one point which
>      >     occurs
>      >     to me is that forest plots are conventional plotted with small
>      >     values of
>      >     standard error at the top.
>      >
>      >     Michael
>      >
>      >     On 19/07/2023 06:07, Gabriel Cotlier via R-sig-meta-analysis
>     wrote:
>      >      > Dear all,
>      >      >
>      >      > I have already posted this question with no response.
>      >      > Maybe this time I am luckier and someone with more
>     knowledge than
>      >     me in the
>      >      > Metafor package can answer me.
>      >      >
>      >      > In a nutshell, what I would like is to be able to produce a
>      >     scatter plot of
>      >      > the observed oucomes or the estimates, in my case Fisher's
>     z for
>      >     the x axis
>      >      > and the standard error in the y axis, with the standard error
>      >     (SE) the
>      >      > same as it appears when running the funnel() function for a
>      >     funnel plot
>      >      > with the model (without moderators) as the input argument.
>      >     Actually, it is
>      >      > a funnel plot without the background of the funnel
>     distribution
>      >     but just
>      >      > the scatter of points, that is suppressing the funnel
>     distribution on
>      >      > the background.
>      >      >
>      >      > I tried to do so in agreement with the definition of SE
>     used for
>      >     the funnel
>      >      > plot in the package Vignette published at Journal of
>     Scientific
>      >     software in
>      >      > page 26:
>      >      >
>      >      > "*For models without moderators, the figure shows the observed
>      >     outcomes on
>      >      > the horizontal axis against their corresponding standard
>     errors
>      >     (i.e., the
>      >      > square root of the sampling variances) on the vertical axis. A
>      >     vertical
>      >      > line indicates the **estimate based on the model. A pseudo
>     confidence
>      >      > interval region is drawn around this value with bounds
>     equal to
>      >     ±1.96 · SE,
>      >      > where SE is the standard error value from the vertical axis.*"
>      >      >
>      >      >
>      >      > I tried to reproduce the vertical axis (y) using the
>     square root
>      >     of the
>      >      > sampling variable, but the result was an upside down
>     scaling of the
>      >      > observed outcomes or estimates on a different y scale for
>     the x
>      >     ticks. The
>      >      > plot seems to have similarities with the funnel plot from the
>      >     funnel()
>      >      > function, but it is not exactly the same without the
>     background
>      >     of the
>      >      > funnel distribution graphic. Maybe the problem could be
>     that in the
>      >      > funnel() function, contrary to my simple attempt to imitate it
>      >     with the
>      >      > square root of the sampling variable, the pseudo confidence
>      >     interval is
>      >      > estimated for each value? Could this be the reason?
>      >      >
>      >      >
>      >      > If so, how could I reproduce the funnel () function plot
>     without
>      >     the funnel
>      >      > distribution graphic in the background and just the
>     scattering of the
>      >      > points using the same pseudo-confidence interval?
>      >      >
>      >      >
>      >      > Thanks a lot for your help and assistance.
>      >      >
>      >      > Kind regards,
>      >      >
>      >      > Gabriel
>      >      >
>      >      >       [[alternative HTML version deleted]]
>      >      >
>      >      > _______________________________________________
>      >      > R-sig-meta-analysis mailing list @
>      > R-sig-meta-analysis using r-project.org
>     <mailto:R-sig-meta-analysis using r-project.org>
>      >     <mailto: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>
>      >     <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>
>      >     <http://www.dewey.myzen.co.uk/home.html
>     <http://www.dewey.myzen.co.uk/home.html>>
>      >
>      >
>      >
>     <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>
> 
>     -- 
>     Michael
>     http://www.dewey.myzen.co.uk/home.html
>     <http://www.dewey.myzen.co.uk/home.html>
> 

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



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