Oct 18th 10:14pm update: Live Blog: 2010 Calgary mayoral election (Naheed Nenshi, Ric McIver, Barb Higgins)
***
I usually just ignore TV news programs’ online polls. But when I saw a TV show this morning publishing their own online poll results of the 2010 Calgary voters’ preferences for mayor, I decided enough is enough. I have to write something about TV news polls in general.
Why TV news polls are unreliable?
TV news polls ask their viewers to go online or call a number to register their views, and then they report these results as “news”. A lot of TV stations have done this so the disease is wide spread! :)
Sure, don’t get me wrong, these polls are nice tools to engage viewers and make them feel their voices are heard but they are also useless as a scientific and rigorous polling tool.
“Reliable polls” share one common criteria that is easy to appreciate. The individuals selected for the polls have to be fair, that is everyone has to have the same “chance” of being selected, in other words “random” (a geeky explanation at wikipedia). And TV news polls, by its nature, suffer from self-selection bias. What the heck is self-selection bias? Here is a simple explanation, any poll you can volunteer yourself and get your friends to vote is not “fair” because we can “stack the deck”. Here is better and more rigorous explanation,
“In statistics, self-selection bias arises in any situation in which individuals select themselves into a group, causing a biased sample with nonprobability sampling.”
So I hope I managed to shine some light on the unreliability of TV news polls.
Note: Polls done by large polling firms are usually done in a more rigorous manner. But it never hurts to, once in a while, question the methodology of the polls and try to find out the questions being asked and the conclusions being drawn from the questions. Think for yourself and don’t blindly trust.
Subscribe to RSS Feed
Subscribe by email
