Cosma Shalizi's "null model of scientific inquiry" [1] seems relevant here. Basically it's a model where all published findings are due to the file drawer effec

Cosma Shalizi's "null model of scientific inquiry" [1] seems relevant here. Basically it's a model where all published findings are due to the file drawer effect [2]. In such a model, scientific results will indeed have a "half-life" until they are overturned. [1] http://masi.cscs.lsa.umich.edu/~crshalizi/weblog/698.html http://masi.cscs.lsa.umich.edu/~crshalizi/weblog/698.html [2] http://en.wikipedia.org/wiki/Publication_bias http://en.wikipedia.org/wiki/Publication_bias