Sampling of all kinds remains extensively used in the social sciences, including economics and
accounting. But achieving the degree of sample randomness sufficient to allow the precise
application of the probability calculus remains elusive if not impossible. We believe that the most
important action that researchers can employ is to assure that important results in terms of economic
impact are supported by replicating the research. Whether such replications will find acceptance in
the various forms of publication remains to be seen. In the interim, we encourage the use of the
SSRN and other Internet options.