LA Times Poll Kerry up 51-44 (without Nader) Kerry up 48-42 (with Nader)

[Smash255]

http://www.latimes.com/news/custom/timespoll/la-na-poll10jun10,1,1874410.story?coll=la-home-headlines

[Umengus]

is it a good poll institute?

[© tweed]

The Vorlon will tell us

[CARLHAYDEN]

The Vorlon will tell us

Don't know about Vorlon, but there are a number of problems with the poll.

First, its registered voters, not likely voters.

Second, the LATimes polls have a history of tilting to the left.

Third, summer polls typically give the Democrats a couple of points higher than non-summer polls (for a couple of reasons).

[Lunar]

is it a good poll institute?

I kind of doubt it.  I originally thought the 51-44 numbers would be for California when I saw the topic.

[Fmr. Gov. NickG]

It's kind of strange they would have Kerry up 7 nationally but tie or trailing in Ohio and Wisconsin...but I guess not impossible given each poll's MoE.  It would be interesting to know what Party ID weight they used, though.  (I can't get into the website)

I wouldn't dismiss the poll just because it is registered, as opposed to likely, voters.  There are lots of registered voter polls at this stage of the game, and they have not shown a consistant bias toward Kerry or Bush.

[agcatter]

Like the rest of us, I await the Vorlon analysis.

[ThePrezMex]

What is weird is the way the three states they polled (apart from the national poll) behave. In Missouri Bush is 48% against 37% for Kerry and 5% for Nader. While bush is +2 in Wisconsin and -3% in Ohio.
Yet, Bush approval numbers are the following:
Overall:
National 51 app - 47 disapprove
Wis        51-45
OH         48-48
MO         49-45

Iraq:
National 44-55
Wis        45-51
OH         43-52
MO         44-50

Terrorism:
National 54-42
Wis        55-40
OH         55-39
MO         52-41

Economy:
National 43-54
Wis        47-47
OH         40-55
MO         42-49


With those numbers, how can Bush be +11 in MO but behind in OH and +2 in Wis?

[JNB]

   Some of the polls have very odd results. I am not as much surprised about MO, since it has become more southren in nature in the last few years, especially its rural areas. OH is where I expect it to be, the WI result is off from other polls that show Kerry more than 5 points ahead.

  Again, I need to see in these polls the Liberal, Morderate, Conservative breakdown, the ethnic especially white male vote and the gender gap.

[CARLHAYDEN]

It's kind of strange they would have Kerry up 7 nationally but tie or trailing in Ohio and Wisconsin...but I guess not impossible given each poll's MoE.  It would be interesting to know what Party ID weight they used, though.  (I can't get into the website)

I wouldn't dismiss the poll just because it is registered, as opposed to likely, voters.  There are lots of registered voter polls at this stage of the game, and they have not shown a consistant bias toward Kerry or Bush.

Good analysis.

I can see Bush being up a couple of points in Ohio, and down a couple of points in Wisconsin, but the inversion in the LATimes polls just doesn't make sense.

[The Vorlon]

is it a good poll institute?

This thing is a ^&^&ing disaster internally

The had a hugely too democratic a sample, and then the weighted it to make it even more democratic.

After weighting the sample has 14% (!!) more democrats that Republicans

Kerry wins democrats  89/8 (+81)
Bush wins Republicans 92/3 (+88)
Bush wins Indys 49/46 (+3)   

   LA Times National Sample            
            
   DEM   GOP   IND   
   0.43   0.29   0.28   
            
Bush   8   92   49   
   0.43   0.29   0.28   
   3.44   26.68   13.72   43.84
            
Kerry   89   3   46   
   0.43   0.29   0.28   
   38.27   0.87   12.88   52.02
            
Rasmussen Weighting            
            
   DEM   GOP   IND   
   0.38   0.35   0.27   
            
Bush   8   92   49   
   0.38   0.35   0.27   
   3.04   32.2   13.23   48.47
            
Kerry   89   3   46   
   0.38   0.35   0.27   
   33.82   1.05   12.42   47.29
            
Zogby Weighting            
            
   DEM   GOP   IND   
   0.37   0.35   0.28   
            
Bush   8   92   49   
   0.37   0.35   0.28   
   2.96   32.2   13.72   48.88
            
Kerry   89   3   46   
   0.37   0.35   0.28   
   32.93   1.05   12.88   46.86

Note enough info on states to do a breakout by party

Interesting notes on states:


   

[agcatter]

Dems up 19 in generic poll?  LOL.

Shades of the California recall.  The LA Times strikes again.

[Filuwaúrdjan]

Generic Ballot Poll=Waste of trees

[The Vorlon]

Generic Ballot Poll=Waste of trees

Wrong - telephone poll => waste of electrons.  

No trees were harmed in the making of this poll

[Filuwaúrdjan]

Generic Ballot Poll=Waste of trees

Wrong - telephone poll => waste of electrons.  

No trees were harmed in the making of this poll

I'm pleased to find out that no trees were harmed

[Fmr. Gov. NickG]

Yuck...that's terrible.

Do you know what the party ID weights were on the OH, MO, and WI polls?

[The Vorlon]

Generic Ballot Poll=Waste of trees

Wrong - telephone poll => waste of electrons.  

No trees were harmed in the making of this poll

I'm pleased to find out that no trees were harmed

Only positive nice thing I can find to say about the entire poll.  

The internals on this thing are an unmitigated disaster.  I would have been embarrased to publish something this clearly ^&^%$ed up.

Every so often you get a bad sample, and then you do the right thing, - throw it away - , and try again. - Yikes.

[The Vorlon]

Yuck...that's terrible.

Do you know what the party ID weights were on the OH, MO, and WI polls?

Not yet, I am working on it.   will post to the "polling" section of the forum when I get a chance. - Should have the .pdfs soon.

Bush +11 in Missouri... now that is funny...  

[Reds4]

Looking at the internals Vorlon has given us and the Congressional Ballot being +19 for Dems, I'd say Bush being down 7 is a pretty good showing in this poll.

[Smash255]

is it a good poll institute?

This thing is a ^&^&ing disaster internally

The had a hugely too democratic a sample, and then the weighted it to make it even more democratic.

After weighting the sample has 14% (!!) more democrats that Republicans

Kerry wins democrats  89/8 (+81)
Bush wins Republicans 92/3 (+88)
Bush wins Indys 49/46 (+3)   

   LA Times National Sample            
            
   DEM   GOP   IND   
   0.43   0.29   0.28   
            
Bush   8   92   49   
   0.43   0.29   0.28   
   3.44   26.68   13.72   43.84
            
Kerry   89   3   46   
   0.43   0.29   0.28   
   38.27   0.87   12.88   52.02
            
Rasmussen Weighting            
            
   DEM   GOP   IND   
   0.38   0.35   0.27   
            
Bush   8   92   49   
   0.38   0.35   0.27   
   3.04   32.2   13.23   48.47
            
Kerry   89   3   46   
   0.38   0.35   0.27   
   33.82   1.05   12.42   47.29
            
Zogby Weighting            
            
   DEM   GOP   IND   
   0.37   0.35   0.28   
            
Bush   8   92   49   
   0.37   0.35   0.28   
   2.96   32.2   13.72   48.88
            
Kerry   89   3   46   
   0.37   0.35   0.28   
   32.93   1.05   12.88   46.86

Note enough info on states to do a breakout by party

Interesting notes on states:


   

The weight is too high for Democrats, but the breakouts per party make very little sense.  For example Bush is running stong with republicans in most polls, but 92/3 is completley absurd.  So when yo re-weight it using Rasmussen or Zogby's party % it is off a bit because no way is Bush +89 in Republicans pretty much every poll shows it in the +78- +81 range.  By the way where did you get the party breakout info from??

[The Vorlon]

The weight is too high for Democrats, but the breakouts per party make very little sense.  For example Bush is running stong with republicans in most polls, but 92/3 is completley absurd.  So when yo re-weight it using Rasmussen or Zogby's party % it is off a bit because no way is Bush +89 in Republicans pretty much every poll shows it in the +78- +81 range.  By the way where did you get the party breakout info from??

I have posted a .pdf of the full poll breakout.

http://www.uselectionatlas.org/FORUM/index.php?board=5;action=display;threadid=4566

I took the breakouts by party ID from this document.

I actually made a mistake, the poll was 51/44 bush not 52/44 so there are "only" 12.5% more democrats than republicans.

This poll "pushes" the leaners which explains both the 5% undexided and the 92/3 Bush advantage among the GOP.

Kerry at 88/7 reflects a similar push of Dem leaners.

[Smash255]

The weight is too high for Democrats, but the breakouts per party make very little sense.  For example Bush is running stong with republicans in most polls, but 92/3 is completley absurd.  So when yo re-weight it using Rasmussen or Zogby's party % it is off a bit because no way is Bush +89 in Republicans pretty much every poll shows it in the +78- +81 range.  By the way where did you get the party breakout info from??

I have posted a .pdf of the full poll breakout.

http://www.uselectionatlas.org/FORUM/index.php?board=5;action=display;threadid=4566

I took the breakouts by party ID from this document.

I actually made a mistake, the poll was 51/44 bush not 52/44 so there are "only" 12.5% more democrats than republicans.

This poll "pushes" the leaners which explains both the 5% undexided and the 92/3 Bush advantage among the GOP.

Kerry at 88/7 reflects a similar push of Dem leaners.

True, but if the polls is pushing leaners why would Kerry be getting only 3% of the Rep vote.  Most polls, even those that don't push the leaners show him getting 7-8% of that vote.  Anyway point being the poll definatley has too many Democrats, but his Republican support his high in every poll, but this poll is much stronger Republican support than the others so that is a biit off also.  So while you can't take the overall #'s seriously because of the overpolling of Dems its hard to transform these numbers into the Rasmussen or Zogby weighting to get a fair picture because his Republican support looks too high in this poll (92/3??))

[The Vorlon]

The weight is too high for Democrats, but the breakouts per party make very little sense.  For example Bush is running stong with republicans in most polls, but 92/3 is completley absurd.  So when yo re-weight it using Rasmussen or Zogby's party % it is off a bit because no way is Bush +89 in Republicans pretty much every poll shows it in the +78- +81 range.  By the way where did you get the party breakout info from??

I have posted a .pdf of the full poll breakout.

http://www.uselectionatlas.org/FORUM/index.php?board=5;action=display;threadid=4566

I took the breakouts by party ID from this document.

I actually made a mistake, the poll was 51/44 bush not 52/44 so there are "only" 12.5% more democrats than republicans.

This poll "pushes" the leaners which explains both the 5% undexided and the 92/3 Bush advantage among the GOP.

Kerry at 88/7 reflects a similar push of Dem leaners.

True, but if the polls is pushing leaners why would Kerry be getting only 3% of the Rep vote.  Most polls, even those that don't push the leaners show him getting 7-8% of that vote.  Anyway point being the poll definatley has too many Democrats, but his Republican support his high in every poll, but this poll is much stronger Republican support than the others so that is a biit off also.  So while you can't take the overall #'s seriously because of the overpolling of Dems its hard to transform these numbers into the Rasmussen or Zogby weighting to get a fair picture because his Republican support looks too high in this poll (92/3??))

This poll is clearly f*^%ed.. let's just throw it away....

There will be a new poll in the next 20 minutes or so....

PS - Didn't even have to wait 20 minutes...

New Opinion Dynamics Poll -

http://www.uselectionatlas.org/FORUM/index.php?board=5;action=display;threadid=4568&start=0