Tuesday, August 26, 2008
"House Effects" in Polling
Since it's an election year, we'll be talking a lot about the presidential election and its effects on International Relations and American Foreign Policy. To start this discussion, considering this really interesting post about "house effects" in the polling data. Here is a bit more complicated (but also very interesting!!) post on how to think about polling results.
Subscribe to:
Post Comments (Atom)
3 comments:
I think polls are fascinating and fun to watch, even though they mean little considering the Electoral College is still in place.
I'd be interested in seeing if there is a "poll voting effect." That is to say, when potential voters see one candidate leading the other, does the general voting populous become more influenced to vote for the more "popular" candidate?
Furthermore, now that convention season is fully in play, we get to start watching for that "convention bounce." What makes this one so interesting is that the conventions are back-to-back this year, and the gaps between them are so small. This could lead to some interesting poll changes, and perhaps some whacky results in the coming weeks.
I actually heard someone on CNN last night mention that Hurricane Gustav will probably be making landfall (perhaps in the U.S.) while the Republican convention is going on in Minneapolis. If coverage is more focused on the hurricane, is it possible McCain will miss his "bounce" and be handicapped down the stretch? That brings me to my ultimate question: Are convention bounces tangible leads that can be held onto, or are they simply arbitrary leads that dissipate no matter the circumstance? In other words, do the conventions always equal each other out, or can someone hold onto their bounce?
I forgot to post this interesting study by FiveThirtyEight.com. Has some interesting graphs to explain the "convention bounces."
http://www.fivethirtyeight.com/2008/08/what-convention-bounce-looks-like.html
I agreed with what Nic posted on polls. I consider them unreliable informants, because there are so many factors that determine the outcome, but at the same time they are rather important (and at times, fun) to watch. The importance stems from the fact that most Americans get their information from one of the national news networks. In my experience, once a household finds what they believe to be a "reliable" network, they'll pay serious attention to what is being reported without really questioning the reliability (I mean, how else can people proudly proclaim that Fox or CNN is completely unbias?). If said news channel reports on what is the current popular opinion, some will mull this over and say, "Oh yes, this is me as well." I'm not saying that everyone does this, but there must be a significant amount that would or else polling would be completely obsolete.
Post a Comment