[R-meta] Compiling different design in the same met-analysis

Dr. Gerta Rücker ruecker @end|ng |rom |mb|@un|-|re|burg@de
Mon May 3 19:12:56 CEST 2021


Technically, you may conduct a meta-analysis of these two effects and 
obtain something between. The question is how to interpret this.

Best,

Gerta

Am 03.05.2021 um 19:06 schrieb Michael Dewey:
> Yes, exactly. After all if you show two things are different you 
> cannot say they are the same.
>
> Michael
>
> On 03/05/2021 15:59, Gladys Barragan-Jason wrote:
>> Okay, so you can't obtain an overall effect size while controlling 
>> for design?
>>
>> Le lun. 3 mai 2021 à 14:36, Michael Dewey <lists using dewey.myzen.co.uk 
>> <mailto:lists using dewey.myzen.co.uk>> a écrit :
>>
>>     If you include a moderator you do not get an overall effect 
>> because by
>>     including it you are effectively saying there is no single overall
>>     effect.
>>
>>     On 03/05/2021 12:24, Gladys Barragan-Jason wrote:
>>      > Hi Michael,
>>      > Thanks a lot for your quick response. Here are answers and other
>>      > questions below.
>>      >
>>      > Le lun. 3 mai 2021 à 12:41, 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>>> a écrit :
>>      >
>>      >     Dear Gladys
>>      >
>>      >     Comments in-line
>>      >
>>      >     On 03/05/2021 08:13, Gladys Barragan-Jason wrote:
>>      >      > Dear all,
>>      >      >
>>      >      > I am doing a meta-analysis on experimental data on the
>>     impact on
>>      >      > several behavioral interventions including both *pre-post
>>     studies*
>>      >      > (measurement of a trait before and after the intervention)
>>     *and
>>      >     control
>>      >      > vs. treatment studies *(measurement after a control
>>     intervention vs.
>>      >      > measurement after the intervention).
>>      >
>>      >     So the first design just has one group which is measured
>>     twice and the
>>      >     second design has two parallel groups which are each measured
>>     once?
>>      >
>>      >
>>      > Exactly!
>>      >
>>      >
>>      >      > I am interested in the effect size of each type of
>>     intervention on a
>>      >      > specific trait that is common to all studies.
>>      >      >
>>      >      > I would like to know whether it is correct to compile all
>>     studies
>>      >     from
>>      >      > one type of intervention (whatever the type of 
>> comparisons:
>>      >     pre-pot and
>>      >      > control vs. treatment group together) with pre- and
>>     control being as
>>      >      > control and post and treatment being as treatment.
>>      >      >
>>      >
>>      >     I do not think that is going to work well. Although the two
>>     designs
>>      >     could be considered as estimating the same effect their
>>     variability
>>      >     should be very different since in design 1 you have reduced
>>     the effect
>>      >     of individual variability. If you do include both designs 
>> in one
>>      >     analysis I would have thought it mandatory to include a
>>     moderator with
>>      >     two levels for design.
>>      >
>>      > This makes sense. When I do this, I then obtain two effect sizes
>>     (one
>>      > for each design); (in having mods=~design-1). Do you know how I
>>     can get
>>      > the overall effect size?
>>      >
>>      >
>>      >      > I have a second question about how to thank the
>>      >      > 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>>
>>      >      > <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
>>     <mailto:r-sig-meta-analysis using r-project.org>>> list for all their help
>>      >      > (acknowledgements in a paper).
>>      >
>>      >     I would have thought the URL
>>      > 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//>>
>>      >     was better. How to do this will also depend on custom and
>>     practice in
>>      >     your scientific discipline, some journals in mine require
>>     people who
>>      >     are
>>      >     acknowledged to give their consent.
>>      >
>>      >
>>      > Ok! Thanks a lot!
>>      >
>>      > Best,
>>      > Gladys
>>      >
>>      >
>>      >     Michael
>>      >
>>      >      >
>>      >      > Thanks a lot for your help!
>>      >      >
>>      >      > Have a great day,
>>      >      >
>>      >      > Gladys
>>      >      >
>>      >      > --
>>      >      >
>>      >      > ------------------------------------------
>>      >      >
>>      >      > Gladys Barragan-Jason, PhD. ____ Website
>>      >      > <https://sites.google.com/view/gladysbarraganjason/home
>> <https://sites.google.com/view/gladysbarraganjason/home>
>>      >  <https://sites.google.com/view/gladysbarraganjason/home
>> <https://sites.google.com/view/gladysbarraganjason/home>>>
>>      >      >
>>      >      > Station d'Ecologie Théorique et Expérimentale (SETE)____
>>      >      >
>>      >      > CNRS de Moulis____
>>      >      >
>>      >      > image.pngimage.png____
>>      >      >
>>      >      >
>>      >      >
>>      >      >
>>      >      >
>>      > 
>>  <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> 
>>
>>      > 
>>  <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://www.avg.com>
>>     <http://www.avg.com <http://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> 
>>
>>      > 
>>  <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>
>>      >     <mailto:R-sig-meta-analysis using r-project.org
>>     <mailto:R-sig-meta-analysis using r-project.org>>
>>      >      > 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>>
>>      >
>>      >
>>      >
>>      > --
>>      >
>>      > ------------------------------------------
>>      >
>>      > Gladys Barragan-Jason, PhD. ____ Website
>>      > <https://sites.google.com/view/gladysbarraganjason/home
>> <https://sites.google.com/view/gladysbarraganjason/home>>
>>      >
>>      > Station d'Ecologie Théorique et Expérimentale (SETE)____
>>      >
>>      > CNRS de Moulis____
>>      >
>>      > image.pngimage.png____
>>      >
>>      >
>>      >
>>
>>     --     Michael
>>     http://www.dewey.myzen.co.uk/home.html
>>     <http://www.dewey.myzen.co.uk/home.html>
>>
>



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