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 #67: Further response [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 Sat Jun-16-07 09:50 AM
Response to Reply #60
67. Further response
Edited on Sat Jun-16-07 09:50 AM by Febble
If you are correct about this then the NEP seems a great deal more amateurish than I would have thought it was. You keep asserting that there are only 30 urban precincts in the exit poll. I thought there were roughly 1800 precincts in the exit poll or something of that order. 30 seems a tad on the low side.


There were over 1400 precincts in the entire poll. Only a fraction of these were in the National Sample, and of these, only a fraction were in big cities. That brings it down to between 20 and 30.

Your suggestion that the 66% over estimation of turnout is due to sampling error and MOE on the small urban sample is I am afraid hard to believe. It is too big an error even for a deeply flawed poll. For the NEP team to have published these "adjusted" figures for the purpose of serious analysis, knowing that they could be some thoroughly flawed is remarkable and negligent.


I am sorry you find it hard to believe. You would, of course need to know the variance before you could compute the MoE. Bear in mind that some precincts would have had boundary changes between 2000 and 2004, so even the raw data won't be that accurate.

But what you need to bear in mind when you talk about a "deeply flawed poll" is that conducting an exit poll that is not vulnerable to flaws is extremely difficult. Voters are not selected at random. Precincts are pre-selected, also not at random, but stratified in order to be representative, and in such a way that each voter in the country has, theoretically, an equal chance of being selected, if sample sizes are equal at each precinct, as they are intended to be. Large precincts are therefore selected at a higher rate than small precincts. If you vote in a large precinct you have a higher chance that your precinct will be in the poll, but a lower chance of being selected by the pollster.

However, precinct size is determined on the previous election's turnout. If this increases non-uniformly, this will affect the representativeness of your sample. In fact, in general, the precinct selection was fairly good (results in selected precincts tended to be representative of the state as a whole). However, it looks as though the precincts subselected to represent big cities in the National sample were not representative of the urban precincts in general. As there were less than thirty of them this is scarcely surprising, and if variance is turnout change was large, then the MoE will be extremely wide.


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