On May 10, Rasmussen released the results of a poll for U.S. Senator. The respondents were asked, “If the election were held today, would you vote for Republican Thom Tillis or Democrat Kay Hagan?” Despite this heavy-handed attempt to discourage respondents from saying they favor the third candidate, Libertarian Sean Haugh, the results are: Tillis 45%, Hagan 44%, “someone else” (which the respondent had to volunteer) 5%, undecided 7%.
One would think that any pollster who wants accurate results would ask respondents about all three candidates whose names will appear on the ballot. Thanks to PoliticalWire for the link.
Does including all candidates in a poll actually produce “accurate” results?
Ah, Rasmussen, the shill for the two-party system. It seems North Carolina voters are so fed up with both of those parties that a significant percentage of those polled defied the inherent attempt to manipulate their opinion. Maybe one day more voters will once again start showing enough of this quality to elect some third party candidates to Congress.
“Does including all candidates in a poll actually produce “accurate” results?”
Does a horse have a tail?
Does presence of a tail accurately determine that an animal is a horse? Does absence of a tail accurately determine that a horse with no tail is not a horse?
If a pollster reads the name of a candidate does it induce some interviewees to say they support that candidate, even when they won’t actually vote for that candidate? If so, then that makes the result of the poll more inaccurate.
Using Jim Riley’s logic why poll at all?