According to Steve Schmidt, one way in which the Republicans measure Romney's chances of winning election based off of RNC Convention is whether or not the candidate polls above Margin of Error after their speech.
On Wednesday, Steve Schmidt, former senior strategist from the McCain/Palin Campaign told Rachel Maddow that in order to see whether or not Romney is picking up Independent voters and middle of the electorate voters is if he polls above the Margin of Error.
SCHMIDT: Sometime, if Romney is to be elected president of the United States, sometime in the next two months, he`s going to have to pass the president and not look back. And part of what Republicans have to be looking for out of this convention is when he`s done speaking on Thursday night and that polling begins. Friday, Saturday, tough to get samples, you know, for Republicans just over the weekend.
But by Sunday, Monday Republicans want to see Mitt Romney ahead outside the margin of error, you know, for the first time in the race.
Republicans want to see does the Republican argument work. Is there elasticity in it to spread the field for the first time? Is there a response in the middle of the electorate? And that is one of the things that Republicans will be looking at and judging at as we get, you know, through this week and in to the early part of next week.
Today, Saturday, Fox Contributor rightwinged pollster, Scott Rasmussen released his Saturday poll showing Romney is
not leading above the Margin of Error. In fact, Rasmussen has Romney leading in line with their Margin of Error, + / -3%.
I do not trust any of Rasmussen's polling data but since Rasmussen is the only pollster out with data today, I'll use Rasmussen polling data as a litmus to see how Romney's poll numbers have been when compared to Rasmussen's Margin of Error. Rasmussen only has Romney ahead by their Margin of Error one time, June 18-20.
08/08-08/10: Margin of Error +/-3 ... Romney up +2
07/29-97/31: Margin of Error + / -3 ... Romney up +3
07/12-07/14: Margin of Error + / -3 ... Tie
06/18-06/20: Margin of Error + / -3 ... Romney up +4
I think it is very telling that even Scott Rasmussen is not even able to manipulate his data points to give Romney a lead above the margin of error post RNC Convention.
Prior to the convention, Rasmussen claimed they had Obama was at 47 and Romney was at 45 -- but that data was still within the Margin of Error -- so why did Scott Rasmussen have Obama up by 2 prior to RNC convention?
It's my opinion that Rasmussen had Obama up by 2, at 47, at the beginning of the RNC Convention because not many people understand Margin of Error and therefore, Scott thinks not many people will realize that giving Mitt 47 after the RNC Convention really means, there was no movement in Romney's poll numbers because the data was still within Rasmussen's Margin of Error. In other words, Scott Rasmussen could have manipulated his data points to deceive.
According to Steve Schmidt, we can all look at the polls coming out today, tomorrow and Monday and if Romney is ahead above the Margin of Error, that would be a good indicator for Romney's chance of winning the election -- However, if Romney cannot poll above the Margin of Error, that would be a bad indicator for Romney as it would mean he does not have a good chance of winning the election.
So far, it is not looking good for Romney. Even Rasmussen can't manipulate his data enough to get Romney above their Margin of Error.
I wonder what the poll numbers would indicate for Romney if Romney where to give "the People" details on his Tax Policy, Foreign Policy and why he, Romney, invests in Foreign Countries but not in America?
Refresher course in basis Statistics (nothing to do with content of Diary)
The Margin of Error is a calculated number where the number of people polled is in the denominator. Since the number of people polled is in the denominator, the more people polled, the smaller the Margin of Error will be. Conversely, the fewer people polled, the larger the Margin of Error.
As and example, using a Margin of Error of 4:
A Margin of Error suppose to be a confidence indicator which means that if you asked a question from a poll 100 times, 95 of those times the percentage of people giving a particular answer would be within +/-4 points of the percentage who gave that same answer in this poll.
There are inherent problems with all polls and statistical data. One problem is "who" was polled, what were the demographics and things of that nature.
To me, the worst problem with statistical analysis is that certain data points can be tossed out and the people looking at the data often times do not know data has been tossed out nor do they know why certain data was disregarded.
For instance, in a political poll, if 1000 people were surveyed but on 900 data points were used, then it would be nice to know why 100 data points were tossed out.
-- End of basic statistics refresher coarse.