National Poll...on Polling

News Story:

A new national poll by the Sacred Heart University Polling Institute investigates the relationship between media, academic and private pollsters and Americans.

Interestingly, the Sacred Heart University researchers found that 10.9% of Americans who have participated in a poll said they have lied to a pollster.

Nearly one-third of all Americans surveyed, 30.9% said they have participated in a political, policy or issue poll in the past. Nearly three-quarters suggested that polls, generally, are very or somewhat accurate and very or somewhat believable – 71.0% and 72.0% respectively.

Over three-quarters, 77.1%, suggested that candidates and elected officials who say they don’t listen to polls are lying. Another 79.3% agreed that elected officials should see polls as an important source of information on where the public stands on issues.

On bias, more survey respondents saw poll question wording as biased or leading (33.5%) than those who saw questions as fair and unbiased (25.0%). The remainder were unsure or neutral.

Two-thirds, 67.6%, agreed that polls are usually accurate and that it is how the media report results that is misleading or biased.

“They complain about our calls, admit to an occasional lie, and question the fairness of wording, but 67.1% said they are interested in poll results and 60.3% said they enjoy seeing how they fit in with the rest of the nation,” according to Jerry C. Lindsley, director of the Sacred Heart University Polling Institute.

EXPERT AVAILABLE FOR COMMENT

  • Jerry Lindsley, director, Sacred Heart University Polling Institute

    To speak with this expert, please contact Funda Alp at 203-396-8241 or alpf@sacredheart.edu.

How the Poll Was Conducted
The Sacred Heart University Polling Institute completed 800 interviews with residents nationwide between March 3-15, 2008. The sample was generated proportional to population contribution in all 50 states. Statistically, a sample of 800 completed telephone interviews represents a margin for error of +/-3.5% at a 95% confidence level.