Dumbf*ck du Jour

L.A. Times writer Lynn Smith

In an article attributed to her that was published in the 3 March 2006 issue of the Washington Post Express, there is this little gem:

The number of channels viewed does not increase substantially until the number of available channels rises above 110, Nielson researchers said. Homes receiving more than 40 channels watched 12 to 20 of them; in homes with an average of 158 channels, 20 of them are actually watched.

WTF? I'm not sure what I even got told, let alone what it is supposed to illustrate. Does the subsample of 40+–channel–homes watching 12-to-20 of them include the subsample of mean-158–channel–homes watching 20 of them? Is the 12-to-20 range the range of channels watched at the 40+–channel homes, or is it the range of means for various subsets of the 40+–channel homes? And assuming, just for fun, that the quotation means exactly what it says (whatever that may be), how exactly do the 20 channels watched in 158-channel homes represent a 'substantial increase' over the 20 from "12 to 20" channels watched in 40+–channel homes?