[FOUNTAIN]Using and abusing polls
Published: 06 Jan. 2003, 00:37
The problem with the Literary Digest's 1936 poll was its sample selection, even though the sample was large. It selected people from such sources as the yellow pages, car registration lists and its own subscriber list. The group was all middle- or upper-class people and supporters of the Republican Party.
In the 1948 U.S. election, the major public opinion research institutes predicted that the Republican, Thomas Dewey, would win the presidency over Harry Truman. Truman beat Dewey by 5 percentage points in the actual election. A photo of Mr. Truman holding the Chicago Tribune that reported a victory by Mr. Dewey is one of the most famous photos in the history of American politics.
In the 2000 U.S. presidential election, the heads of broadcasting networks and news agencies that predicted the Florida election outcome incorrectly were summoned to the Congress for a hearing. Recent surveys about who would win the governors' races in Georgia and Illinois and the senate election in Colorado were also incorrect. Telephone polling is getting questionable now because people increasingly avoid answering their telephones; they screen calls with caller ID or answering machines. But polling is the only way to know what voter sentiment really is. That is why Roh Moo-hyun, the Millennium Democratic Party presidential candidate, and Chung Mong-joon, the National Unity 21 presidential candidate, want to use polls to consolidate their candidacies. But they are deliberately playing with the polls; watching them is like reading Aesop's fable in which a fox put some soup in a very shallow dish for his friend the stork and the stork in return put soup in a long-necked jar for the fox.
The writer is a deputy foreign news editor of the JoongAng Ilbo.
by Noh Jae-hyun
with the Korea JoongAng Daily
To write comments, please log in to one of the accounts.
Standards Board Policy (0/250자)