Polling Results:how polls work - Tilted Forum Project Discussion Community

tecoyah · 09-07-2004, 05:52 AM

Okay....here we go, experiment successful.
The point of this experiment was to define how polling is manipulated.

Media report#1. Republican Poll

President Bush currently holds 68 % approval raitings in the tecoyah polls

Media report#2. Non Partisan Poll

President Bush currently holds 33 % approval ratings in the tecoyah polls

Media report#3. Democratic Poll

President Bush currently holds 7 % approval ratings in the tecoyah polls

As you can see the wording of the questions will create the results of ANY poll. All three of these questions could be used to guage the popularity of the President, and would be reported as a valid result. These results were of no suprise to me, and I hope were at least somewhat enlightening.

Thank you for participating.

Zeld2.0 · 09-07-2004, 10:54 AM

Which is why poll information without actually giving you information on what questions were asked and how random the sample size was is hard to prove anything...

livingfossil · 09-07-2004, 02:23 PM

Question: "The screaming terrorists are running, screaming, at your home... with bombs. Would you rather have GWB with a gun defending you, or JK with a white handkerchief?"

or

"GWB is a man who has bankrupted several businesses. Will you entrust the economy to him? Or JK?"

polls, polls, polls. Most (in not all) my lit-theory training has been to decry statistical data. Sometimes I think those crazies in the Literature departments have a point.

Blackthorn · 09-07-2004, 02:53 PM

There are three types of lies: Lies, damn lies, and statistics.

smooth · 09-07-2004, 04:18 PM

tecoyah,

your polls were interesting. Without doing a validity check on them, however, they don't indicate anything significant.

How do you know they asked the same question? I mean, that's your point: that they asked the same question, worded differently, which resulted in different responses for the same question.

Most polls and statistical tools utilize multiple questions, asked in various ways, in order to ascertain a particular theme. For example, it's likely that all three of your questions would be on a poll and then the data analyzed to see what people were really answering. Of course, they wouldn't even be used until they had been checked for validity.

The linguistics student does have a point, however, in that studying language can give deeper insight and the way things are worded should be suspect. But I don't know of any university educated people (during which time students must complete some statistical courses) who believe that polls or statistical data are outright nonsense.

09-07-2004, 05:52 AM	#1 (permalink)
tecoyah Illusionary	Polling Results:how polls work Okay....here we go, experiment successful. The point of this experiment was to define how polling is manipulated. Media report#1. Republican Poll President Bush currently holds 68 % approval raitings in the tecoyah polls Media report#2. Non Partisan Poll President Bush currently holds 33 % approval ratings in the tecoyah polls Media report#3. Democratic Poll President Bush currently holds 7 % approval ratings in the tecoyah polls As you can see the wording of the questions will create the results of ANY poll. All three of these questions could be used to guage the popularity of the President, and would be reported as a valid result. These results were of no suprise to me, and I hope were at least somewhat enlightening. Thank you for participating. __________________ Holding onto anger is like grasping a hot coal with the intent of throwing it at someone else; you are the one who gets burned. - Buddha

09-07-2004, 02:23 PM	#3 (permalink)
livingfossil Tilted Location: Chicago	Question: "The screaming terrorists are running, screaming, at your home... with bombs. Would you rather have GWB with a gun defending you, or JK with a white handkerchief?" or "GWB is a man who has bankrupted several businesses. Will you entrust the economy to him? Or JK?" polls, polls, polls. Most (in not all) my lit-theory training has been to decry statistical data. Sometimes I think those crazies in the Literature departments have a point. __________________ Never anything witty.

09-07-2004, 02:53 PM	#4 (permalink)
Blackthorn Beware the Mad Irish Location: Wish I was on the N17...	There are three types of lies: Lies, damn lies, and statistics. __________________ What are you willing to give up in order to get what you want?

09-07-2004, 04:18 PM	#5 (permalink)
smooth Junkie Location: Right here	tecoyah, your polls were interesting. Without doing a validity check on them, however, they don't indicate anything significant. How do you know they asked the same question? I mean, that's your point: that they asked the same question, worded differently, which resulted in different responses for the same question. Most polls and statistical tools utilize multiple questions, asked in various ways, in order to ascertain a particular theme. For example, it's likely that all three of your questions would be on a poll and then the data analyzed to see what people were really answering. Of course, they wouldn't even be used until they had been checked for validity. The linguistics student does have a point, however, in that studying language can give deeper insight and the way things are worded should be suspect. But I don't know of any university educated people (during which time students must complete some statistical courses) who believe that polls or statistical data are outright nonsense. __________________ "The theory of a free press is that truth will emerge from free discussion, not that it will be presented perfectly and instantly in any one account." -- Walter Lippmann "You measure democracy by the freedom it gives its dissidents, not the freedom it gives its assimilated conformists." -- Abbie Hoffman

09-07-2004, 10:54 AM	#2 (permalink)
Zeld2.0 Cherry-pickin' devil's advocate Location: Los Angeles	Which is why poll information without actually giving you information on what questions were asked and how random the sample size was is hard to prove anything...