Talk:Straw poll

Iowa Straw Poll
This page has existed for years, yet it's still a stub. Could someone explain how the Iowa straw polls work? Anyone else got anything to add? 71.178.1.32 03:39, 30 March 2007 (UTC)


 * See the Ames Straw Poll article. &mdash;Lowellian (reply) 01:28, 23 July 2007 (UTC)

I've been traveling to straw polls in Iowa, NH, and SC for the Republican's John Cox and now Daniel Gilbert. Strawpolls are not regulated except by the local county GOP party that sets the rules. More often than not, these are fundraising strawpolls, meaning that ANYONE can vote and any number of tickets can be purchased meaning that any number of votes can be purchased. A couple of weeks ago I went to the Palmetto Family Council Strawpoll at the Columbia Convention Center, to benefit the Palmetto Family Council. Mike Huckabee got first, Ron Paul got 2nd. There was a disclaimer in the ticket purchase form and on their website that proper ID would have to be given so tickets could be restricted to one ticket per individual. In reality this provision was never enforced, we were free to buy as many tickets as we pleased for $50 each. In addition, 90% of the total attendance were candidates plus their staff and invites, only 10% were the general public so the winner was simply the one who bought the most tickets. In buying these tickets, campaigns also tried to not overspend, Huckabee was simply more willing to buy out the straw poll for their advantage than Ron Paul since they are making a guestimate on how much to spend to win.

A couple of months ago in Manchester, NH, Ron Paul won solely because he had more supporters on hand, who collectively pooled their cash (tickets sold for cash only) and bought their victory for $3800. Total ticket sales were more than 8 times the actual attendance.

Other strawpolls however are strictly controlled by the county GOP to restrict voting to state residents and even registered voters. These are the only reliable polls to watch. The last strawpoll near Georgetown, SC in the Waccammaw School Auditorium was restricted to SC residents.

If it is a fundraiser strawpoll, it should be completely ignored since the winner with the most available cash can buy out their victory, for sale to the highest buyer. Wtp4prez 13:59, 9 October 2007 (UTC)
 * Observations (and I've heard Cox speak in person and respect his intentions and sentiments): (1) Straw polls generally, but not always, take ordinary steps to prevent duplicate voting; when they don't (as is usually the case online), they should not be called straw polls in my opinion. General elections also take strong steps to prevent duplicate voting, but cannot prevent all attempts to obtain multiple voter registration cards. So duplicate voting cannot be perfectly enforced, it can only be reasonably prevented and morally deprecated. (2) You didn't at any point allege that any person physically submitted multiple votes in any straw poll, just that persons bought multiple tickets in excess of attendance. By contrast, I know of a "poll" in Olathe, KS, where multiple voting was explicitly permitted. I understand a Fredhead won by purchasing 50 Thompson votes for $1 each, without the benefit of 49 like-minded friends; only about 100 total votes were purchased in the whole poll. (3) Note that Paul's campaign generally doesn't buy these tickets: as you admit, his supporters do. If a campaign pays $50 a ticket, and has warm-bodied voters and have not prevented them from voting against their own sponsors, the fact that the campaign has this money and is willing to spend it on this result and can find the voters is, in fact, significant and exactly what straw polls measure. Even more significant if supporters spend $3800 to win a straw poll: when have you ever heard of that kind of grassroots excitement? (4) By the way, Romney, who is owed $17 million by his campaign, paid over $2000 per vote at the Iowa Straw Poll, partly because a myriad of $35 tickets went unvoted, and because of busing and advertising. That is, the same money wins a whole poll in Paul supporters' hands, but buys exactly two votes in Romney's hands. (5) Finally, I agree that most straw polls are not well-documented and this certainly hampers reporting and collation. But there is no warrant for completely ignoring results that testify to some campaign or supporter persistence and passion. John J. Bulten 20:02, 25 October 2007 (UTC)

Ecclesiastical Use
During the Papal election, there are no straw polls. Once in the conclave all votes are binding. Rather, straw, wet or dry, was burned along with the ballot papers, to produce black (no Pope) or white smoke (a Pope has been elected) to indicate the result of that round of voting. Straw has now apparently been replaced by chemicals. —Preceding unsigned comment added by 203.214.147.34 (talk) 02:30, 27 October 2007 (UTC)

Interactive Dialogues
“Straw polls provide important interactive dialogue”. Is there any kind of dialogue other than an interactive one? Paul Magnussen (talk) 03:43, 12 July 2011 (UTC)

Radio programme called "Straw Poll"
Wasn't there at one time a BBC Radio Four programme called "Straw Poll", which was put on at the same time as Any Questions? in the days when Any Questions took a summer break? If any one remember, and considers it notable enough to have an article in Wikipedia, it could get an article in Wikipedia, and there could then be a disambiguation tag at the top of the page (or a tag saying "For the radio programme, see..." and then a link to the article). ACEOREVIVED (talk) 15:48, 9 May 2012 (UTC)

Statistical errors
"A margin of error is intrinsic in any subset polling method, and is a mathematical function of the difference in size between the subset and the larger population;"

The margin of error includes several sources of error, one of which is sampling error. The sampling error is related to the sample size n and the population size N through the finite population correction (fpc), which is a factor of (1 - n/N). There is no (n - N) term in the sampling error and it is not generally present in other sources of error.

"sampling error is constant across different poll methods with the same sample sizes."

No. In general, there's no reason for it to be anywhere close.

Blaise (talk) 16:50, 31 August 2014 (UTC)