Political Polling

Who Has The Best Poll?

Use all of this Information to Rank at least 5 Polls before a single, or many, elections.

Polling Research
  • LV stands for Likely voters, while RV stands for Registered Voters
  • LV is considered more accurate due to the fact that a formula is used in LV to find who is actually going to physically go vote, while RV is just if they are registered to vote, which anyone with a driver’s license who is above 18 in most states is registered
  • IVR , or interactive voice response, has been proven to be just as effective as traditional calling, except when used as a poll of a poll. Then the IVR polls are much worse than the traditional, which is believed to be due to the polling of a poll. Traditional polls do better in an already used pool of poll takers ( Though it is disconcerting to see pollsters use other polls as a safety check the accuracy is effective in IVR polls)This information has been given by Robo-Polls: TakingCuesfromTraditionalSources”?
  • Avoid push polling by looking for sample questions from the poll analysis (Don’t Use a Pll with Attacking Questions)
  • DPV, or double preference voting, allows polling to have 200% of the data for whom people support, which after formulated show support for third parties out of undecided, making the accuracy of undecided voter polling more accurate. But DPV polling is not accepted well because the two parties polling tends to gain more accurate result in our very party split country
  • According to Nate Silver recency, sample size, and pollster rating affect the accuracy of polls. The more recent the poll the more accurate and the larger the sample size the better.
  • I discovered that locality of a polling agency also leads to better results.
  1. Go to www.realclearpolitics.com and click on the “polls” tab
  2. Search for a race that you would like to rank polls on (after you click the polls tab there will be a search tab in the top right corner of the screen where you can choose a state and the type of race to follow)
  3. Two weeks prior to voting day look up polling and click on the polling agency’s name to print off a polling report (When you click on the race you will see a box with a list of polls in it, that is where you click on the agency’s name to bring up the polling report)
  4. Use the polling reports and use the background information above to decide which poll is more accurate based on the sample size, recency, likability, and locality.
  5. Rank your polls a day or two before the race
  6. After the race get results online, Google works fantastic for this, and see how close your rankings were.
  7. Then write a conclusion statement to tell why your rankings were off. This could be due to a last minute news article about a candidate, a small sample size, the poll is old, or the race just didn’t run out like it was supposed to.
  8. Turn in your statement with what you learned to Mr. Pethan and get lots of points!