ShadesOfGrey
Golden Member
- Jun 28, 2005
- 1,523
- 0
- 0
Originally posted by: Czar
Originally posted by: ShadesOfGrey
Originally posted by: Czar
Originally posted by: ShadesOfGrey
Originally posted by: Bowfinger
Probably. What would you do differently in the second thread, however? We have an OP asserting "cBS" skewed the results of a poll, not based on his understanding of sound statistical science, but because of his partisan emotional need to reject bad news about the Bush administration. I did try briefly to refocus the thread on the poll's results, but most of the conversation has focused on the OP's assertion of impropriety. In essence, it's another variation of science vs. faith arguments.Originally posted by: Czar
After reading two threads this morning, both long, both turning out very stupid, are we just arguing because we like arguing?
threads in question
http://forums.anandtech.com/messageview...atid=52&threadid=1726785&enterthread=y
http://forums.anandtech.com/messageview...atid=52&threadid=1727173&enterthread=y
No, you tried to divert the thread from it's obvious focus on bad sampling. The thread was never about the individual supposed "results" of the polling questions.
The amount of willing ignorance in that thread, especially by you, is amazing. But yes, I agree with the OP that people like you are just arguing for the sake of arguing. You have had nothing to offer that thread and have only served to divert it.
please dont... I read the thread, you are just as guilty of arguing for arguings sake as he is, even when totalcommand posted their polling methood
Well this thread is about my thread so according to Bowfinger logic you want to talk about the substance of my thread here. So tough.
BTW, their claimed polling method has been exposed - which is what my thread is about. They took a political topic, did a poll, then weighted the results and IGNORED party affilitation in doing so. They have a bad sample - period.
ok so to sidetrack the original topic of this thread
people who vote are not the same as those asked in the poll. The poll went to everyone registered or not and when they decided to add a bigger sample they decided so because of age, race, sex, education and so on so it fits the census over the whole country.
not how they voted last time or who they support.
"ok so to sidetrack the original topic of this thread" Bingo. Now you see what Bowfinger was doing in my thread? I did to this thread what he did to mine. I did it to this thread to show him(and you) what he did.
Political affiliation has alot to do with people's stances and opinions. To use an over sample of one(or rather drastinc undersample) you may still have the proper generic demographics but if you are taking a political poll - you can't ignore political affiliation - which cBS seems to have done.