Skip to main content

View Diary: The simple change in wording that shows why you should ignore polls on the NSA leak (140 comments)

Comment Preferences

  •  As someone who worked for a prominent market (7+ / 0-)

    research firm, I know how to write a poll, for a price, in such a way to give you the exact results you want which you can then release to the public (and it will be a legitimate poll that no one will question.)

    Here's how it works: We would, for some clients, do private in-house polling that was not meant for public release. In this polling we would do, say, eight separate polls, each one asking the same questions, but each in a slightly different way.

    You would be amazed how just slightly different wording would end up with much different polling numbers.

    When we figured out which way to word the question gave the client the closest to the result he wanted, we would then do a big, legitimate, nationwide poll asking the question in just the right way. This is the poll that would be released to the public.

    I never trust a poll completely. And I always look very closely at the wording. After a while you kind of learn key words that can skew any poll your way.

    •  What I do is look not only at the wording (1+ / 0-)
      Recommended by:
      Williston Barrett

      but I try to look at many different polls on the same subject, over time. Which is why I trust polling on Social Security--because I've looked at polls over decades.

      Your point here is very well taken.

      Defending the theft of our freedom by the government is not a legitimate difference of opinion on a political matter -- it is a deeply un-American attitude that deserves nothing but scorn and derision.--Dallasdoc

      by SouthernLiberalinMD on Thu Jun 13, 2013 at 08:08:04 AM PDT

      [ Parent ]

Subscribe or Donate to support Daily Kos.

Click here for the mobile view of the site