2016 Postmortem

exboyfil

(17,863 posts) Thu Oct 18, 2012, 07:06 PM Oct 2012

How good are Susquehanna Polling and Research ?

They have Romney up by 4% in Pennsylvaina.

http://washingtonexaminer.com/poll-shows-romney-leading-in-blue-pennsylvania/article/2511153

I know the source is dicey, but I was just wondering.

6 replies

= new reply since forum marked as read

Highlight:

How good are Susquehanna Polling and Research ? (Original Post) exboyfil Oct 2012 OP

From Wikipedia: tarheelsunc Oct 2012 #1

So do they give the candidates two sets of numbers? exboyfil Oct 2012 #2

Does this poll really exist? Mass Oct 2012 #3

Why Wouldn't They Release It To The Press So It Can Be Included In The Polling Averages? DemocratSinceBirth Oct 2012 #4

The fieldwork is from ages ago Ebadlun Oct 2012 #5

538 says Obama has a 91% chance of winning n/t ejbr Oct 2012 #6

tarheelsunc

(2,117 posts)

1. From Wikipedia:

Reply to exboyfil (Original post)

Thu Oct 18, 2012, 07:11 PM

Oct 2012

The firm specializes in polling services for Republican candidates, trade groups, businesses and lobbying firms in Pennsylvania, New Jersey, Delaware, and Maryland.

exboyfil

(17,863 posts)

2. So do they give the candidates two sets of numbers?

Reply to tarheelsunc (Reply #1)

Thu Oct 18, 2012, 07:13 PM

Oct 2012

The feel good numbers for the masses, and the real numbers to know how bad your getting your butt kicked. That is what I suspected.

Mass

(27,315 posts)

3. Does this poll really exist?

Reply to exboyfil (Original post)

Thu Oct 18, 2012, 07:14 PM

Oct 2012

Anyway, this seems to be an internal GOP poll, so I would not spend too much time on it.

DemocratSinceBirth

(99,710 posts)

4. Why Wouldn't They Release It To The Press So It Can Be Included In The Polling Averages?

Reply to Mass (Reply #3)

Thu Oct 18, 2012, 07:22 PM

Oct 2012

I think your answer lies here:

Susquehanna has traditionally shown a much tighter race between Obama and Romney than other polls, in part because it weighs its results by party registration. Firms that don't do this tend to over-sample Democrats.

Weighting results in Pennsylvania is particularly important, Lee said, "because we know with a pretty good degree of certainty how many registered Republicans and Democrats are going to show up" on Election Day.

Their weighting has to be so biased as to be laughable.

Ebadlun

(336 posts)

5. The fieldwork is from ages ago

Reply to exboyfil (Original post)

Thu Oct 18, 2012, 08:16 PM

Oct 2012

Susquehanna's automated poll of 1,376 likely voters was taken between Oct. 11 and 13, before the second presidential debate Tuesday that many saw as a comeback for Obama since his Oct. 3 showdown with Romney.

God knows why it's only showing up now, but it's an interesting archeological find from the distant era of the Romney bump.

ejbr

(5,856 posts)

6. 538 says Obama has a 91% chance of winning n/t

Reply to exboyfil (Original post)

Thu Oct 18, 2012, 08:55 PM

Oct 2012

Reply to this discussion