Comments on: Wednesday open thread http://sistertoldjah.com/archives/2008/10/22/wednesday-open-thread-27/ Don't dis or dismiss this miss! Fri, 10 Feb 2012 13:18:54 +0000 hourly 1 http://wordpress.org/?v=3.2.1 By: 2Hotel9 http://sistertoldjah.com/archives/2008/10/22/wednesday-open-thread-27/comment-page-1/#comment-751153 2Hotel9 Wed, 22 Oct 2008 22:42:02 +0000 http://sistertoldjah.com/?p=7882#comment-751153 Sis, the only poll that matters happens in November. Sis, the only poll that matters happens in November.

]]> By: Rich Horton http://sistertoldjah.com/archives/2008/10/22/wednesday-open-thread-27/comment-page-1/#comment-751151 Rich Horton Wed, 22 Oct 2008 17:30:24 +0000 http://sistertoldjah.com/?p=7882#comment-751151 The question is, if news organizations can be "in the bag" and biased in favor of one candidate, why shouldn't that extend to the "polling" being commissioned by the very same media? There are many different ways polling can be shaded. You could riddle it with "priming" problems meant to hurt one of the candidates being asked about. You could "weight" the populations in a misleading fashion. There are a good half dozen other things one could do to get the results one wants. "Ah!" the objection is raised, "but then the polling outfits would loose credibility?" But "credibility" with whom? For starters the media outlets who want the bad polling are not going to expose it. Secondly, the issues involved in statistical sampling are difficult to explain to large segments of the population - if they cant "get" margin of error terms, how are they going to "get" weighted samples? Third, it isn't as if polling outfits are selling material to the public at large. Their client base is ONLY the self-same biased media. So they have nothing to lose on that front either. Much of the polling has been suspect, and was even during the Democratic primary. The fact that major polling services have "new" models that just so happen to show Obama with a huge lead compared to the older models which show a much closer contest, is being sold as merely "happenstance." I for one am not retarded enough to believe it. The question is, if news organizations can be “in the bag” and biased in favor of one candidate, why shouldn’t that extend to the “polling” being commissioned by the very same media? There are many different ways polling can be shaded. You could riddle it with “priming” problems meant to hurt one of the candidates being asked about. You could “weight” the populations in a misleading fashion. There are a good half dozen other things one could do to get the results one wants.

“Ah!” the objection is raised, “but then the polling outfits would loose credibility?”

But “credibility” with whom? For starters the media outlets who want the bad polling are not going to expose it. Secondly, the issues involved in statistical sampling are difficult to explain to large segments of the population – if they cant “get” margin of error terms, how are they going to “get” weighted samples? Third, it isn’t as if polling outfits are selling material to the public at large. Their client base is ONLY the self-same biased media. So they have nothing to lose on that front either.

Much of the polling has been suspect, and was even during the Democratic primary. The fact that major polling services have “new” models that just so happen to show Obama with a huge lead compared to the older models which show a much closer contest, is being sold as merely “happenstance.”

I for one am not retarded enough to believe it.

]]>