Wednesday, October 1, 2008

Good example of a bad poll

From Media Matters:

Summary: A Time/SRBI poll asked likely voters with an "unfavorable opinion" of Sen. Barack Obama to respond to various "reasons that voters give us for having an unfavorable opinion of Barack Obama," which included: "He's really a Muslim and not a Christian"; "He's an elitist who doesn't understand the needs of ordinary people"; and, "He's not as patriotic as he should be."


These are leading, biased questions, and rather useless because of it. Anyone trained in this stuff will tell you that you won't get honest results when you plant the options in the heads of the people taking the survey.

As well, Media Matters notes that similar questions were not asked of McCain.

Somewhat humorously, this reminds me of the questionnaire being circulated in Lebanon about math.

It's like saying: "So-and-so is a giant loser. Agree or disagree?" rather than saying either "what is your opinion of so-and-so" or "rate the success of so-and-so on a scale of 1 to 10." The language has to be neutral - results will change depending on how the question is phrased.

1 comments:

Anonymous said...

That is a poll designed to answer a question where the pollster believes he already knows the answer. In other words, he is planning on writing an article about people who don't like Obama because he is an elitist, or Muslim, or whatever, and now will have a "poll" to back up his article. It is crap, but it probably works for some people. I would argue that people who would believe a "poll" like that should not be allowed to vote, but that is just me!

 
Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.