Could anyone point me to a good citation making the case (ideally, supported empirically) that availability of free, open-source software is a necessary precondition for disseminating new methods, especially trial designs? My target problem domain is of course oncology dose finding, but a generic argument would work just fine.
FYI: I’m looking in particular for a citation to back up the claim in next-to-last paragraph here.
You might want to look up the book by Ben Klemens:
Math You Can’t Use: patents, copyright, and software Brookings Institution Press, November 2005.
He has a number of interesting papers from an economic perspective related to statistical models. In particular:
Modelling with Data: Tools and Techniques for Scientific Computing (A full text pdf is available on the author’s page – Modeling with Data )
A Useful Algebra of Statistical Models, which appears to be a technical report for the U.S. Census bureau
His book describes the use of his Apophenia statistical library written in C and is based upon the GNU Scientific Library.