PA PrimD: Zogby: Clinton's PA Lead Back Up to 4%

[Queen Mum Inks.LWC]

New Poll: Pennsylvania President by Zogby on 2008-04-17

Summary:

Clinton:	47%
Obama:	43%
Other:	2%
Undecided:	8%

Poll Source URL: Full Poll Details

Pennsylvania

 4-16/17

 4-15/16

 

Clinton

 47%

 45%

 

Obama

 43%

 44%

 

Someone else

 2%

 3%

 

Not sure

 8%

 9%

[Queen Mum Inks.LWC]

Last poll less than 24 hours earlier

Clinton 47% (+2)
Obama 43% (-1)

I would try to read into this as being maybe a minor Clinton debate bump, but...Zogby.

I say bad 1st sample.

[Queen Mum Inks.LWC]

Last poll less than 24 hours earlier

Clinton 47% (+2)
Obama 43% (-1)

I would try to read into this as being maybe a minor Clinton debate bump, but...Zogby.

I wouldn't bother reading much of anything into movement like that anyway.

True - the movement isn't significant, scientifically significant, since it falls w/in the MoE.

[Queen Mum Inks.LWC]

Last poll less than 24 hours earlier

Clinton 47% (+2)
Obama 43% (-1)

I would try to read into this as being maybe a minor Clinton debate bump, but...Zogby.

It's also within the MoE meaning statistically there was no change.

I already said that. 

[Queen Mum Inks.LWC]

Last poll less than 24 hours earlier

Clinton 47% (+2)
Obama 43% (-1)

I would try to read into this as being maybe a minor Clinton debate bump, but...Zogby.

It's also within the MoE meaning statistically there was no change.

I already said that. 

then you're both wrong.

No - it's within the MoE.  Statsitically, there wasn't a change.  You never took a statistics class did you?

[Queen Mum Inks.LWC]

Last poll less than 24 hours earlier

Clinton 47% (+2)
Obama 43% (-1)

I would try to read into this as being maybe a minor Clinton debate bump, but...Zogby.

It's also within the MoE meaning statistically there was no change.

I already said that. 

then you're both wrong.

No - it's within the MoE.  Statsitically, there wasn't a change.  You never took a statistics class did you?

I've taken quite a few stats classes - at the post graduate level, even, and to say there wasn't a change is inaccurate.  Alcon's method of describing it is more accurate, although still bothers me a bit.  Margin of error is not some blanket thing.   A Poll showing one candidate ahead 52-48 with a 5% MOE is NOT a tie.  It is 95% (usually the criteria) likely that the margin is from -1 to +9, and is quite clearly MUCH more likely that the candidate with 52% in the poll is ahead by SOME margin. 

A later poll showing the lead to be 54-46 IS a change.  Do we know FOR SURE that the voters changed?  no  But we do know that is MUCH more likely that the lead grew than that it slipped (although that is POSSIBLE).

semantics it seems, but still it's a pet peeve of mine when commentators call races statistically tied, or say that the polls show NO CHANGE.  anyway, that's all I got on the subject right now.

You're right - in my original post, I said the change was insignificant.  BRTD said there was not change, and I said that I had said the same thing, thinking that he had said there was no significant change.  So, you and Alcon are right, and that's what I said in my first post.