You are viewing an obsolete version of the DU website which is no longer supported by the Administrators. Visit The New DU.
Democratic Underground Latest Greatest Lobby Journals Search Options Help Login
Google

Reply #95: Well, [View All]

Printer-friendly format Printer-friendly format
Printer-friendly format Email this thread to a friend
Printer-friendly format Bookmark this thread
This topic is locked.
Home » Discuss » Topic Forums » Election Reform Donate to DU
Febble Donating Member (1000+ posts) Send PM | Profile | Ignore Sun Jun-17-07 10:37 AM
Response to Reply #80
95. Well,
YOU: This OP otoh is concerned with the wider numbers -
turnout numbers in general - and cross tabs in relation to
them. Most importantly it involves an analysis of the
apparent illogicality of Bush doing better in the cities than
in the rural areas. Contrary to OTOH persistent BS this does
not rely in the first instance on the NEP as its basis but
the actual turnout numbers and returns.

ME: Not BS.  Please read his posts.

If you or someone from the NEP can explain how the tables we
are discussing.

http://www.scoop.co.nz/stories/HL0706/S00165.htm#b

Were actually compiled then that would help in this analysis.
Guesses about how they were compiled and what may make them
wrong are at this point in the discussion misdirection.

ME: Well, I just compiled this one myself, from the National
datafile.  With weights on the proportion of voters from each
category are:

City over 500,000	12.53%
City: 50,000 to 500,000	18.88%
Suburbs	                44.53%
City: 10,000 to 49,999	 7.93%
Rural	                16.13%


With weights off they are:

City over 500,000	10.90%
City: 50,000 to 500,000	20.57%
Suburbs	                45.13%
City: 10,000 to 49,999	 8.48%
Rural	                14.92%


Obviously I do not know exactly how the weights were
computed, but the weights themselves were publicly available.
Here the mean weights by size of place (I log transformed them
before computing the means)

Size of place code	N	Mean    Mean    Std. Dev
                        	log	weight	(log)
City over 500,000	1498	 0.01	1.01	0.50
City: 50,000 to 500,000	2814	-0.20	0.82	0.49
Suburbs	                6193	-0.12	0.89	0.44
City: 10,000 to 49,999	1160	-0.14	0.87	0.38
Rural	                2046	-0.04	0.97	0.45

As you can see, the mean weights for big city respondents is
greater than 1 (log > 0) and the mean weights for
everywhere else was less than 1 (log < 0.  There is
nonetheless considerable variance.

And here are the mean weights for Kerry and Bush voters by
size of place

		               N	Mean	Mean	Std. Dev
Size of place code			log	weight	log
City over 500,000	Kerry	961	-0.07	0.93	0.50
	                Bush	505	 0.19	1.21	0.42
City: 50,000 to 500,000	Kerry	1488	-0.26	0.77	0.45
	                Bush	1276	-0.13	0.88	0.53
Suburbs	                Kerry	3156	-0.20	0.82	0.44
	                Bush	2936	-0.02	0.98	0.42
City: 10,000 to 49,999	Kerry	579	-0.18	0.84	0.37
	                Bush	560	-0.10	0.90	0.38
Rural	                Kerry	876	-0.11	0.89	0.44
	                Bush	1133	 0.03	1.04	0.44


from which you can see that Bush tended to be upweighted more
heavily than Kerry voters, as we know.

YOU: Meanwhile consider this.

You tell us repeatedly that your study of the exit poll data
failed to find any statistical evidence of vote fraud.[/i]

ME: More than that, it revealed a pattern that is difficult
to reconcile with fraud on a scale of millions.  

YOU:As I was thinking about this last night - and partly
prompted by your criticism of the National exit poll urban
sample - I had an idea.

If you are a serious election stealer and you are using the
NEP as your basis for stealing elections then it would be
easily within your power not to steal votes in the Precincts
that were being polled for the NEP.

This would leave no evidence to be discovered through an
analysis of WPE.

Anyway food for thought.

al

ME: Yes, it was something that has been suggested before, but
is contra-indicated by the fact that at state level, the
precinct vote shares in the NEP precincts were a good
estimate of the vote-shares in the entire state.  Check out:

http://www.exit-poll.net/election-night/EvaluationJan192005.pdf

pp 29-30.
Printer Friendly | Permalink | Reply | Top
 

Home » Discuss » Topic Forums » Election Reform Donate to DU

Powered by DCForum+ Version 1.1 Copyright 1997-2002 DCScripts.com
Software has been extensively modified by the DU administrators


Important Notices: By participating on this discussion board, visitors agree to abide by the rules outlined on our Rules page. Messages posted on the Democratic Underground Discussion Forums are the opinions of the individuals who post them, and do not necessarily represent the opinions of Democratic Underground, LLC.

Home  |  Discussion Forums  |  Journals |  Store  |  Donate

About DU  |  Contact Us  |  Privacy Policy

Got a message for Democratic Underground? Click here to send us a message.

© 2001 - 2011 Democratic Underground, LLC