Anyone still up who wants to check some math.

If this question has already by asked and answered could you direct me to the thread. Otherwise, I've been playing with the before and after exit polls from this thread

http://www.democraticunderground.com/discuss/duboard.php?az=view_all&address=203x214619

and for the life of me I can't figure out how only 57 people were added. Can someone else go through and check the number of people added to each subgroup by race between the early and late exit polls? I keep coming up with about 94 people.

Heres what I found but someone should def. double check.


Earlier exit polls with 1963 polled

W. Men 785.2
W. Women 883.35
Non W. Men 117.78
Non W. Women 157.04

Later exit polls with 2020 polled

W. Men 808
W. Women 929.2
Non W. Men 141.4
Non W. Women 161.6

Subtracting the differences I get 93.83 added people.
What am I missing?

A total of 97 added by calculating the changes in the 4 groups.

2020- 1963 =57

So there's a discrepancy of 40 votes.

Nothing to see here, move along.

Kind of figured it had to be something simple...:eyes:

add up to 100%. For what you call early exit polls, the percentages add up to 99%, and for what you call late exit polls, the sum is 101%. There is nothing funny about this -- it is most probably caused by rounding off of the percentages. So 0.01x1963 + 0.01x2020 = 40, thus the difference that you are seeing, i.e. 40+57=97. End of proof. :)

I really don't think it could be a problem with the data at CNN, because the percents I'm coming up with were based on the number of voters ...I'm going to bed now. I'm totally not even able to focus anymore.

websites, which you in turn used to calculate the number of men/women. If you add them up, the sum is not 100%.

101% and 99%. sorry, I just reread my last post. Too vague. I was agreeing with your point. I don't think it was a problem with the CNN data...thanks again for clearing it up. :-)

Hey just realized its the last day of 04 here. :bounce: