Bucky
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:38 PM
Original message |
Are "internal polls" any more reliable than publically released polls? |
|
I sometimes here the phrase "internal polls" used to describe results that a losing candidate relies on. But aren't internal polls just surveys that a campaign takes and doesn't bother to release? Does anyone with experience in campaigns know if these use larger sampling sizes or different methodologies?
I used to work in market research, so I'm pretty familiar with random sampling methodologies and how size of sample can affect margins of error and confidence rates. But is there anything special or unique about internal polls that make them more reliable (or less) than the polls news organizations buy and report on?
|
Drunken Irishman
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:42 PM
Response to Original message |
|
I remember someone posted an Edwards internal poll from Iowa showing him winning Iowa, Clinton coming in 2nd and Obama a distant 3rd.
I'm guessing sometimes they are, but, just like other polls, they can be off by a lot as well.
|
MissMillie
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:43 PM
Response to Original message |
2. I thought they were less reliable (in fact, I think the term I've seen is "unscientific") |
|
It has to do with the fact that you're not getting a random sample.
|
Bucky
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:49 PM
Response to Reply #2 |
3. Why would they do a poll without drawing a representative sample? |
|
So there really is a difference in the methodology then? How is the sampling done for internal polls?
|
MissMillie
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:50 PM
Response to Reply #3 |
4. I'm sorry... I read your OP incorrectly |
|
I thought it said "internet polls".
my bad
|
grantcart
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:50 PM
Response to Original message |
5. yes if they are doing daily canvassing because they are looking |
|
at a larger base of data and watching it trend on a daily basis.
|
geek tragedy
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:51 PM
Response to Original message |
6. They're better, but only leaked selectively, i.e. cherry-picked. |
|
If you see internals released, it means that it's because the campaign wants you to see them. If their polling for 7 days is awful and one day is great, they'll leak the one day's polling.
|
NoBorders
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:53 PM
Response to Original message |
7. I've wondered that also |
|
I wonder if they dont' adjust their samples to check different scenarios - add in less likely voters, more youth, women, etc... to give them an idea what would happen 'if'...
They're also probably getting some intelligence 'on the ground' from their field and GOTV operations that polling companies don't have, so they may be able to see somewhat solid anecdotal stuff that isn't reflected in the polls we see.
|
Occam Bandage
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:54 PM
Response to Original message |
8. They're often more up-to-date, but other than that, not really. |
NJSecularist
(1000+ posts)
Send PM |
Profile |
Ignore
|
Tue Feb-05-08 03:55 PM
Response to Original message |
9. Internal polls usually give candidates a better idea of how they are doing by each district |
|
While public polls are more general, and don't poll specific districts and counties. They poll a whole state.
|
DU
AdBot (1000+ posts) |
Fri Apr 26th 2024, 01:29 PM
Response to Original message |