posted on Feb, 14 2012 @ 08:46 AM
reply to post by seachange
Originally posted by seachange
I stopped watching after a couple minutes do to extreme wrongness by the Youtube author. You can easily predict the results on an election based on a
sample size of 2,300. This is common sense. All the major polling companies regularly make statements about 300 million Americans based on sample
sizes of rarely over 3,000 people. This is how statistics works.
There is no hoax, just people making wrong assumptions about polling that are false. The margin of error isn't based on the available maximum sample
size. Everyone in statistics knows that. Just because you can get more samples doesn't mean you need to get more samples. The video author of the OP
is simply wrong.
There is no hoax.
Originally posted by darkstar57
I hate being a skunk at this garden party, but I once taught statistics. I also agree with the sentiments and point of view of the majority of ATSrs.
But the following is true.
2391 is adequate to give a plus or minus 3 percent at 95 pc confidence to a percentage of voters that voted for say, the Newt.
IF the exit polls were properly conducted and sampled.
check the link for the ifs that have to be answered to get an accurate and representative result.
Your reactions were my initial reactions as well. However you both also understand how awesome statistics are at manipulating our views and
understandings (this should be an emphasized part of every statistics course, but it was glossed over in all of mine). Especially how drastically
"projection" statistics swing with very subtle changes in source data and data collection methods.
According to the actual vote data they were showing, they had more actual votes than exit poll votes. Yet they projected based on the exit poll
numbers rather than their actual numbers. I can't tell for sure, but even if you merge the two it looks like their projection is still off from the
actual numbers they had in front of them. It's right there.
Someone may want to argue that the exit polls were from a more complete sample set, but I will simply argue back that when you get to pick your
samples (however "random"), the results shouldn't be used for anything. Only the actual votes and only project a winner when you have enough there is
no mathematical way someone else could win.
Discourage voters by convincing them their side has already lost. Very old method going back as far as voting and a basic tenet of war. The best way
to win the war is to convince the enemy you already have. Then if discovered later claim a mistake by the network(s) and apologize. Rinse and
edit on 2012/2/14 by ErgoTheConfusion because: (no reason given)