Monday, October 3, 2011
How to read a poll
In its traditional media role of scorekeeper, we will see more and more polls reported between now and the beginning of caucuses and primaries in Jan. These days, it seems there's a poll for everything. In this episode of Congress.org's D.C. Decoder, host Craig Crawford shows you how to tell if a poll's results are legitimate.
www.rollcall.com/cqpolitics/dc_decoder/48423-1.html
Subscribe to:
Post Comments (Atom)
2 comments:
From this video I learned that not all polls are reliable and you need to check to see if it is a good poll. A reliable poll will list at the bottom the methods of how the data was collected. A good poll does not need to survey thousands of people. Rather only a few hundred people who are diverse enough to represent the whole population will create a good poll. Also, you need to check the margin of error on a poll because this will let you know how accurate the statistics are. Finally, you need to look at the wording of the questions and the order they were asked in. This will tell you if the poll was biased and led people to respond a certain way or if the poll is unbiased.
- Emma Woods
I learned that certain polls can be biased and don't give the opinion of the population as a whole (like a TV host giving a poll that only his viewers will respond to). There are different types of polls that are used by people with biased opinions that collect data from people who more than likely share the same opinion, such as internet polls that do not include certain data. I learned that "methodology", the fine print at the bottom of the poll, will tell you how the poll gathers its information and the "margin of error" is the amount calculated in a poll that is likely to be misinformation. I also learned that biased polls will word their questions differently to lead you to answer those questions a certain way.
Post a Comment