Posted by Kati Togliatti
The paper “Statistical Analyses and Reproducible Research” by Robert Gentleman and Duncan Temple Lang discusses the principles underlying reproducible research. The authors define reproducible research as research papers that have accompanying software so that the reader can directly reproduce results. They talk about the use of a compendium which is something that combines test, data, and software into one package that can be distributed. The compendium is composed of dynamic documents, which can be used by different people in varying ways. In class this is currently being taught and it seems like github or dropbox could be considered “compendiums” because multiple users can access the data and use it in different ways. In my own personal work I am planning on making a folder with all of my research data in it so that one day someone else can hopefully follow the work I’ve done to continue it.