Reply #4: Lampley may have his heart in the right place

[dsc]

but his math is not. Here is just one example of a fairly large mistake.

Byron York's most recent refutation of my charge that irregularities in the 2004 Presidential election demand criminal investigation cites quotes from the report of Edison/Mitofsky, the two-company partnership which provided exit polls to the major television networks, on the vast discrepancies between those polls and the official results of the election. The report, which Mr. York has helpfully highlighted in his second post and which runs to about eighty pages, essentially offered the conclusion that an five-and-a-half point gap between final poll numbers and the national popular vote tabulation-- a variance more than four times the statistical margin for error of 1.3%-- can be attributed to shy Republicans. The Washington Post summarized the conclusion: "procedural problems compounded by the refusal of large numbers of Republican voters to be surveyed led to inflated estimates of support

He is off by a factor of two here. MOE's work in both directions, not only one. Thus when they predict that Bush and Kerry are going to be tied but end up with them apart by around 5%, they are only off by a 2.5 MOE (2.5 each way). He has made other similar mistakes in other posts. I applaud his liberalism, I don't applaud his sloppiness.