[R-meta] The trim and fill method for rma.mv function

[Michael Dewey]

Dear Rafael

Removing outliers without an external justification is usually not 
recommended as it (a) is data-dependent, (b) they might be the most 
interesting observations, (c) it is better to alter the model to fit the 
data, not alter the data to fit the model.

Michael

On 31/10/2017 22:22, Rafael Rios wrote:
> Dear Wolfgang,
> 
> Thank you. I tried the approach that you suggested using the standard error
> as moderator, and I found a bias in my data set. I removed outliers (i.e.
> values outside the 95% confidence interval of the funnel plot), but I still
> found a p<0.05 for the intercept. Then, I decided to removed values outside
> 90% confidence interval using a contour-enhanced funnel plot and do the
> meta-analyses. Do you agree with my approach or recommend a different one?
> Best wishes,
> 
> Rafael R. Moura.
> *scientia amabilis*
> 
> Doutorando da Pós-graduação em Ecologia e Conservação de Recursos Naturais
> Universidade Federal de Uberlândia, Uberlândia, MG, Brasil
> 
> ORCID: http://orcid.org/0000-0002-7911-4734
> Currículo Lattes:
> http://buscatextual.cnpq.br/buscatextual/visualizacv.do?id=K4244908A8
> Research Gate: https://www.researchgate.net/profile/Rafael_Rios_Moura2
> 
> 
> <http://buscatextual.cnpq.br/buscatextual/visualizacv.do?id=K4244908A8>
> 
> 2017-10-31 8:17 GMT-02:00 Viechtbauer Wolfgang (SP) <
> wolfgang.viechtbauer at maastrichtuniversity.nl>:
> 
>> Dear Rafael,
>>
>> The trim and fill method has not been generalized to more complex models,
>> so this is why this is not available in metafor.
>>
>> It is relatively straightforward though to extend the 'regression test' to
>> more complex models. Simply include some measure of precision of the
>> estimates (e.g., the standard error, the (inverse) sample size) as a
>> predictor/moderator into the model.
>>
>> Best,
>> Wolfgang
>>
>> --
>> Wolfgang Viechtbauer, Ph.D., Statistician | Department of Psychiatry and
>> Neuropsychology | Maastricht University | P.O. Box 616 (VIJV1) | 6200 MD
>> Maastricht, The Netherlands | +31 (43) 388-4170 | http://www.wvbauer.com
>>
>> -----Original Message-----
>> From: Rafael Rios [mailto:biorafaelrm at gmail.com]
>> Sent: Tuesday, 17 October, 2017 20:52
>> To: r-sig-meta-analysis at r-project.org; Viechtbauer Wolfgang (SP)
>> Subject: The trim and fill method for rma.mv function
>>
>> Dear All,
>>
>> I conducted a mixed-effects model using rma.mv function using metafor
>> package. Now, I want to use trimfill function, but it is not applyied to
>> rma.mv function. Do you know any alternative to conduct this analysis and
>> graphical vizualization?
>>
>> Best wishes,
>>
>> Rafael R. Moura.
>> scientia amabilis
>>
>> Doutorando da Pós-graduação em Ecologia e Conservação de Recursos Naturais
>> Universidade Federal de Uberlândia, Uberlândia, MG, Brasil
>>
>> ORCID: http://orcid.org/0000-0002-7911-4734
>> Currículo Lattes: http://buscatextual.cnpq.br/buscatextual/
>> visualizacv.do?id=K4244908A8
>> Research Gate: https://www.researchgate.net/profile/Rafael_Rios_Moura2
>>
> 
> 	[[alternative HTML version deleted]]
> 
> _______________________________________________
> R-sig-meta-analysis mailing list
> R-sig-meta-analysis at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-meta-analysis
> 

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