FOAS_logo_and_title

Philosophy

The mission of FOAS is to promote free software, open access publishing, and reproducible research in statistics. We understand free software to be as defined by the Free Software Foundation: possible to run for any purpose, available in source code form, and free to modify and/or redistribute. Open access forums for papers are, for us, defined as those that allow free access to all readers with an internet connection, and that invite contributions from authors without requiring the payment of fees.

Most `open access' journals are only open to those authors able to pay fees (often hundreds or thousands of dollars) to submit papers. The result is scientific discourse in which the only parties allowed to speak are those who can pay, or who are specially waived in (e.g., due to citizenship in a developing country). FOAS seeks to support forums for the publication of results in statistics that are open access for both readers and authors. The Journal of Statistical Software and The R Journal are both open access in this full sense.

Currently, the prime example of the free software we seek to support is R, the lingua franca of computational statistics. The source code of R is freely available online, and possible to modify, extend and distribute under the terms of the GNU General Public License. The base R distribution is also easily extensible with user-contributed packages, many thousands of which have been added to the Comprehensive R Archive Network.

FOAS also advocates for reproducible research, which entails the publication of code and data used to reach conclusions in the literature. In the same spirit that requires that the results of a scientific experiment be published alongside a description that allows another scientist to repeat the study, so the reproducible research movement advocates for the publication of code and data that allows the conclusions of a study to be independently validated. Ideally, the code is also possible to run on software for which the source code is available, so that other researchers can examine the computations performed in the analysis at any level desired. The Journal of Statistical Software distributes code along with all of its articles, making it a good example of a journal that fosters reproducible research.