Reply #43: You say... [View All]

<snip>

"MITOFSKY ET ALL ADMIT THAT VOTES WERE MIXED
WITH THE POLLS. THE EVIDENCE SHOWS IT BEGAN AROUND 1:00am
NOV.3. WILL THE NAYSAYS SAY NAY ABOUT THAT? OF COURSE THEY WILL.
THEY WOULD ALSO SAY THAT NEW YEAR'S DIDN'T HAPPEN."

</snip>

The first presidential exit poll in 1967 mixed exit poll results with actual votes after the polls closed. The same thing happened this election, and in 2008, the exit poll numbers will again be 'mixed' with actual votes as precincts close. THIS IS THE WAY EXIT POLLS HAVE ALWAYS BEEN DONE, AND ALWAYS WILL BE DONE.

You've been told this before, but you always reply with some comment about low post counts. Please educate yourself before doing more damage. Thanks.

<snip>

How did Bush get 63.6% of the final 623 out of the 3168 total
who responded to the dual question of HOW or IF they voted in
2000? Bush only had 48.1% of the FIRST 2545 polled.

</snip>

Well, it all depends upon the makeup of the final 623 of 3168. If the democrat/republican ratio was the same as the first 2545, then you may have something here. What was the party breakdown of the final 623 ? You are assuming the 623 is a random sample of the 3168. Is it a random sample ? If you can't answer these questions, then you have nothing.

Now, I expect your reply to me to be no different from your reply to everyone else that criticizes you (low post counts, partisan hack, freeper, whatever.) Such vacuous arguments may work in a DU forum, but they won't work in the real world. The real world is where this is supposed to be going, right ? We hope we can get into court or in front of congress, right ?

In the real world, people who are experts in their respective field will be called to testify. They will destroy an analysis such as the one you provide here. You need to stop focusing your argument on DU members and begin refining the argument you intend to use when defending your work against the actual experts called by the other side. You can't use your usual arguments (low post counts, partisan hack, freeper, whatever) on the exit poll experts called by the other side. These experts will say things like:

"It's not possible to use the raw exit poll numbers for any type of useful statistical analysis." You may think the experts are wrong in this regard, but you must explain why the experts are wrong and you are right. In short, you must actually defend your work. I encourage you to start with me. If you can't convince me, you will never convince an expert.