Substantiating programmer variability

Dickey's critique of the Sackman et al. data is well taken. An alternate data set is presented which substantiates the enormous variability in programmer performance. The opportunity for productivity gains and improved experimental methods in research through reducing the range of this variability remains fertile.