In other words, even if in general models produce a reasonable statistical representation of natural variability, could they be a long way out for the specific case of 1950 - 2010? (realclimate.org)
I've always felt that utilizing statistical representations from which to make investment decisions is fraught with error. (fastgraphs.com)
And what's inside it is not statistical representations of text, but actual representations of knowledge. (slaw.ca)