Praise be to Tavris for her thoughtful article on the misuse of polling data. However, in her desire to discredit pseudo- and unscientific polling she commits the same error she condemns in others--the misinterpretation of polling data. There is no evidence that the Times-Mirror/Gallup series of polls are overly “refined” and “virtually meaningless.” I do not say this as a result of the false consensus effect. As a ‘60s Democrat I would love to eat these words on Nov. 8 as Michael Dukakis pulls off a Harry Truman 40 years later and blows away the polling pundits.
GORDON D. MUNRO