![]()
|
The Conspiracy Letters Saturday, February 01, 2003 FROM A VOLUNTEER POLL-MONITOR You say in "More Krugman Poll-Watching" (1/31/2003), "the 4% difference between 'fair' and 'unfair' is only slightly greater than the 3% margin of error in the poll"But if a poll shows "42% with a 3% margin of error" that means there is a 95% (or some other percentage, but 95% is common) chance that the true result is between 39% and 45%, that is, 3 points above or below. The 38% figure is, then, somewhere between 35% and 41%. A four-point difference is actually within the margin of error. Linda Seebach Posted by Donald L. Luskin at 2:07 PM | link
Friday, January 31, 2003 TAKE THE PLUNGE Your latest volly at Krugman was in line with my teaching so I wrote something out and posted it for public viewing at www-stat.wharton.upenn.edu/~ajw/Plunge.htm.It's a bit technical (intended for my class), but honest and complete. The main point is that there is indeed no plunge, although for a certain point of view it does appear that there has been a big drop lately. Professor Abraham Wyner Department of Statistics Wharton School of Business Posted by Donald L. Luskin at 12:58 PM | link
Sunday, January 26, 2003 QUOTE FILLING You have often written about the journalistic practice of "quote filling," how reporters call pundits to get the quote they want to support an opinion they have already made. Whether reporters are getting more aggressive at this or my own awareness has increased, it is getting harder by the week to avoid stepping into one of those pieces of intellectual poop. In Saturday's Financial Times (January 25, 2003, UK Edition, page 8), James Harding writes:"Unlike last year, Mr Bush faces a re-energised opposition. Nelson Polsby, professor of political science at University of California in Berkeley, notes: 'Things are going to get a lot worse for him in Congress.'"Now, if Mr. Harding had to make a phone call all the way to liberal Berkeley to find someone who agrees with him, his argument is rather weak. I mean, how hard is it to find someone at Berkeley who will have something negative to say about the political future of the current dministration? After all, the same Polsby is quoted elsewhere: "'Who won the 2000 elections?' he asks rhetorically. 'Exactly, the Democrat Al Gore. His party lost slightly this year. It's the normal setback for the party that really gave it its all last time around.' In this reading of the results of mid-term elections the historical trend has not been broken at all."The odds of Polsby telling the FT that things will get a lot worse for Bush are left as an exercise to the astute reader. Sylvain Galineau Posted by Donald L. Luskin at 4:56 PM | link
|