I would like to ask what are the main reasons and the advantages of using R as statistical work rather than using any other software which can help the statistician for doing his work ?
Gordon is right, free is probably the best reason. In addition, the R software program has an outstanding support network of researchers that have shared their research experience, tips, and tricks through blog posts and discussion boards. The analysis section of the metafor package (http://www.metafor-project.org/doku.php/analyses) also has examples of nearly every type of statical analysis you would want to run for a meta-analysis using R, along with references to cite in your own publications.
Good Luck!
Mike
Dear sami,
You should know that there is something is unique in R,It isn't a stat programme however,it is programming language and software environment for statistical computing and graphics. I could tell you what is unique may be network meta analysis or hasse diagram but i need to confirm it is painful at first.
Kind regards,
Dear Sami,
I've written a couple programs that help people run a meta-analysis and are all based on metafor and R. MAVIS is a Shiny app and R package for running a meta-analysis in a web browser kylehamilton.net/shiny/MAVIS/ also MAJOR is a module for an SPSS replacement program called Jamovi https://www.jamovi.org/news/2017/11/23/new-modules.html
Feel free to try them out and let me know what you think I'm always looking for new features to add to my software.
Kyle
Dear Sami,
R has significant advantages over other softwares used in meta-analysis as following:
- R is a free, open source, and powerful statistical environment, it has more than 20 meta-analytic packages on CRAN.
- Has tools for meta-regression, Bayesian meta-analysis, and multivariate
meta-analyses.
- By R, you can change the appearance of plots (forest and funnel plots) as you like.
- Can make different types of funnel plots as standard, trim and fill or contour-enhanced funnel plots.
- R generates L'Abbé plot by metafor package which is used to assess heterogeneity.
- Can assess publication bias by Begg's and Egger's tests.
- R can conduct multiple treatments and indirect treatment analyses.
- Not only can do meta-analysis but also systematic reveiw. R can search, store, and import data from NCBI databases by RISmed package.
I would say that RevMan and other point-and-click programs are suitable for basic meta-analysis and for newcomer to meta-analysis, but if you have the flexibility to learn something new, change what things look, and use more advanced statistical programs, then R or STATA are the best choices.
I found a few advantages to use metafor package of R.
1. You can produce nice Funnel plot for illustrating publication Bias. There are at least three types you can generate by using R - a) General funnel plot, b) Contour-enhanced funnel plot, c) Trim & fill funnel plot.
2. For epidemiological study-based meta-analysis, you can generate L'Abbé Plot which is quite unique way to represent heterogeneity.
3. You can calculate Begg's and Egger's regression tests for quantitative analysis of heterogeneity calculating p values.
However, for generating Forest plots, I would go for RevMan.
Thanks
Good luck!
- Badges
- Science method
- Similar topics
- Mathematics
- Statistics
- Statistical Techniques
- Meta-Analysis