Viechtbauer Wolfgang (SP)
wolfgang.viechtbauer at maastrichtuniversity.nl
Sun Jun 11 13:11:42 CEST 2017
(This is in part just a test and also a first post to r-sig-meta-analysis)
I think it's fair to say that R has become one of the primary tools for conducting meta-analyses. The meta-analysis task view now lists close to 80 different packages that provide some kind of functionality for conducting meta-analyses with R:
Accordingly, questions and discussions on how to conduct meta-analyses with R are on the rise. Such questions have been addressed in part on some of the other R-related mailing lists (e.g., r-help, r-sig-mixed-models), on several Stack Exchange sites (in particular, Cross Validated and Stack Overflow), and, in my own personal experience, via e-mail. There are some problems with these outlets:
1) Having meta-analysis related questions/discussions scattered around various lists makes it more difficult to find answers that have already been provided. It also makes it difficult for people interested in 'meta-analysis + R' to participate in the discussion -- one shouldn't have to sign up for (and monitor) a number of different lists (where the vast majority of questions are NOT related to meta-analysis) to participate in the discussion.
2) Not everybody likes the 'gamification' of the Q&A process that is inherent to the Stack Exchange websites. Also, many questions related to 'meta-analysis + R' are partly statistical, partly about coding itself. That creates a bit of tension for the Stack Exchange sites -- Cross Validated is meant for stats related questions, Stack Overflow for coding related issues, but many questions fall in-between.
3) Personal emails are problematic, as they leave no public trace/record of the answers provided, potentially leading to a lot of duplicated efforts. Also, the task of answering questions is then placed on a single individual.
The r-sig-meta-analysis mailing list is meant to address these issues: A single list dedicated to questions/discussions around 'meta-analysis + R', with a single archive (https://stat.ethz.ch/pipermail/r-sig-meta-analysis/) that one can easily browse through or search, without any gamification aspect, appropriate for stats and programming related questions, and in a public forum so that more than just a single person can provide answers and participate in the discussion.
A few tips/notes:
Before participating in any of the R-related mailing lists, please read the Posting Guide first: https://www.r-project.org/posting-guide.html
It is typically much easier to provide an answer when a question comes with a clear reproducible example. See the following link for some tips on how to create a good reproducible example: https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example
By default, there are no search capabilities for the archives. However, Google and other search engines should start to crawl the archives eventually, at which point one can just use regular search engines for searching. For example, add "site:https://stat.ethz.ch/pipermail/r-sig-meta-analysis/" (without quotes) to your Google search to restrict your search to the r-sig-meta-analysis mailing list archives. Note that search engines do not update their records daily, so you may still want to browse the archives manually to find the most recent posts.
I hope that this mailing list will not purely be a place for package authors to answer questions from users running into difficulties, but more generally a place to discuss anything related to 'meta-analysis + R'.
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
More information about the R-sig-meta-analysis