Meta-Analysis

The problem: There are a lot of figures in published papers in the scholarly literature, like the below, from (Attwood et. al. 2012)): At some point, a scientist wants to ask a question for which they can synthesize the knowledge on that question by collecting data from the published literature. This often requires something like the following workflow: Search for relevant papers (e.g., via Google Scholar). Collect the papers. Decide which are appropriate for inclusion. Collect data from the figures using software on a native application. Examples include GraphClick and ImageJ. Proof data. Analyze data & publish paper. This workflow needs revamping, particularly in step number 3 - collecting the data. This data remains private, moving from one closed source (original publication) to another (personal computer). We can surely do better. ...

—UPDATE: I am now using code formatting from gist.github, so I replaced the old prettyR code (sorry guys). The github way is much easier and prettier. I hope readers like the change. I wrote earlier about some code I wrote for running Phylometa (software to do phylogenetic meta-analysis) from R. I have been concerned about what exactly is the right penalty for including phylogeny in a meta-analysis. E.g.: AIC is calculated from Q in Phylometa, and Q increases with tree size. ...

A while back I posted some messy code to run Phylometa from R, especially useful for processing the output data from Phylometa which is not easily done. The code is still quite messy, but it should work now. I have run the code with tens of different data sets and phylogenies so it should work. I fixed errors when parentheses came up against numbers in the output, and other things. You can use the code for up to 4 levels of your grouping variable. In addition, there are some lines of code to plot the effect sizes with confidence intervals, comparing random and fixed effects models and phylogenetic and traditional models. ...