SurveyUSA 14-State Poll Release (user search)
       |           

Welcome, Guest. Please login or register.
Did you miss your activation email?
May 15, 2024, 03:08:08 PM
News: Election Simulator 2.0 Released. Senate/Gubernatorial maps, proportional electoral votes, and more - Read more

  Talk Elections
  Election Archive
  Election Archive
  2008 Elections
  2008 U.S. Presidential General Election Polls
  SurveyUSA 14-State Poll Release (search mode)
Pages: [1]
Author Topic: SurveyUSA 14-State Poll Release  (Read 7750 times)
Small Business Owner of Any Repute
Mr. Moderate
Atlas Icon
*****
Posts: 13,431
United States


WWW
« on: April 19, 2008, 01:42:57 PM »

Young people in Kentucky must really hate Obama.  Clinton also does better among African-Americans, so I don't know what happened there.

Also, why can't Obama seem to do well in Massachusetts?  This still baffles me.

They really like the Clintons there... I truly believe that has a lot to do with it. 

....and racism.

I haven't seen any polls out of Rhode Island, but is there any similar effect noticable?

No, Obama does fine in Rhode Island.  Mainly because Deval Patrick isn't governor there.
Logged
Small Business Owner of Any Repute
Mr. Moderate
Atlas Icon
*****
Posts: 13,431
United States


WWW
« Reply #1 on: April 19, 2008, 01:50:23 PM »

How much do these stations pay for these polls?

In my past experience, a decent poll with a MoE of about 4 or 5 will run you $2,000 if done by live, non-accented operators.  SUSA's costs may be lower, because they use automated response.  Basically...it's not expensive to do.

I can make a call a minute, or 180 calls in three hours between 5:30 pm and 8:30 pm, or 540 calls over three nights. Assuming a 20% response rate, six people can make enough calls. Assuming a payment rate of $10 an hour over 5 people over 9 hours each, that is $450. I can fit all of them into the basement of my house. So what is to stop me from picking 3,240 random names out of the White pages for a particular locality, and spend $450 to conduct my own poll?

Or rather, what is to stop the Atlas forum from collectively conducting a poll? If we had 20 volunteers, each would only need to make (600 x 5 / 20) = 150 calls, which can be made in the space of three hours. All we would need is for one person to cut the list. Then there is a simple formula to calculate margin of error. If our poll comes out as better than a reputable pollster, we have just established ourselves, gentlemen!

There's nothing to stop you (us?) persay, but there is more to polling than just asking questions.  It is almost a science to properly rotate questions and ask them in the right order, and it crucial to find people who have no attachment to the results of the poll reading the choices to the people.  Bias is really easy to pollute a poll, and it takes a good pollster to filter that stuff out best as possible.
Logged
Pages: [1]  
Jump to:  


Login with username, password and session length

Terms of Service - DMCA Agent and Policy - Privacy Policy and Cookies

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines

Page created in 0.025 seconds with 14 queries.