PPP-IA: Braley +1

[SPC]

While this is a partisan-affiliated poll, I have not noticed a substantive difference between their public and sponsored polls this cycle. While this adds slightly more credence to the idea that Ernst has lost ground, the fact that every other poll shows her up by 1-4 points demonstrates that she still has the edge.

[SPC]

Oh, Republicans here.

Poll shows Ernst up six: "We're going to net eight seats!"
Poll shows Braley up one: "Nah, this can't be tied."

Both polls deviated from the consensus of pollsters. It would be foolish to evaluate a race based on individual polls as opposed to the aggregate picture that all the legitimate polls tell.

[SPC]

The numbers are not good for Ernst. In other polls, she went from 6% leads to just 2%, so that indicates that Braley is closing.

In reality, Ernst went from -4 to tied in the Loras poll, tied to +2 in the marist poll, and tied to +4 in the Suffolk poll.


Someone's closing all right.

Wrong as usual. There were at least two polls that showed Ernst up six, but that lead collapsed when both pollsters polled again (DMR and Quinnipiac). Further more, Ernst is hitting a ceiling of 47-48, which is about what she will get on election day.

Perhaps that earlier Selzer poll and Quinnipiac were outliers, because I can't imagine things have improved this quickly for Bruce Braley.

Don't let basic statistics get in the way of the narrative!

I should note that of the three polls showing Ernst's position worsening, PPP's is the only one that seems that it could potentially be indicative of a trend, if only because their previous numbers were right in line with the consensus to begin with. However, the fact that several other pollsters have now shown Ernst with a slight lead indicates that PPP's poll being an outlier is just as much of a possibility.

[SPC]

The numbers are not good for Ernst. In other polls, she went from 6% leads to just 2%, so that indicates that Braley is closing.

In reality, Ernst went from -4 to tied in the Loras poll, tied to +2 in the marist poll, and tied to +4 in the Suffolk poll.


Someone's closing all right.

Wrong as usual. There were at least two polls that showed Ernst up six, but that lead collapsed when both pollsters polled again (DMR and Quinnipiac). Further more, Ernst is hitting a ceiling of 47-48, which is about what she will get on election day.

Perhaps that earlier Selzer poll and Quinnipiac were outliers, because I can't imagine things have improved this quickly for Bruce Braley.

Don't let basic statistics get in the way of the narrative!

What is your point? Go sit down, please.

The idea that a company that previously had a poll deviating from the consensus by several points is now posting numbers consistent with other pollsters is not evidence of "momentum" for Braley; more likely it is attributable to random chance. As I stated earlier, the only definitive evidence we have so far of a Braley comeback is a partisan poll from PPP, and until other polls start to show similar movement away from the consensus of a slight Ernst lead, PPP's result should not be taken at face value.

[SPC]

The numbers are not good for Ernst. In other polls, she went from 6% leads to just 2%, so that indicates that Braley is closing.

In reality, Ernst went from -4 to tied in the Loras poll, tied to +2 in the marist poll, and tied to +4 in the Suffolk poll.


Someone's closing all right.

Wrong as usual. There were at least two polls that showed Ernst up six, but that lead collapsed when both pollsters polled again (DMR and Quinnipiac). Further more, Ernst is hitting a ceiling of 47-48, which is about what she will get on election day.

Perhaps that earlier Selzer poll and Quinnipiac were outliers, because I can't imagine things have improved this quickly for Bruce Braley.

Don't let basic statistics get in the way of the narrative!

What is your point? Go sit down, please.

The idea that a company that previously had a poll deviating from the consensus by several points is now posting numbers consistent with other pollsters is not evidence of "momentum" for Braley; more likely it is attributable to random chance. As I stated earlier, the only definitive evidence we have so far of a Braley comeback is a partisan poll from PPP, and until other polls start to show similar movement away from the consensus of a slight Ernst lead, PPP's result should not be taken at face value.

You're right, PPP is lying and faking polls for Braley. As always, you people are 100% right!

Welcome to the ignore list.

[SPC]

ITT: people who don't understand MoE.

The point of a polling average is to reduce the standard error, as a polling average effectively increases the sample size. If we were to treat the polls taken in October by PPP, Suffolk, Quinnipiac, Selzer, Loras, and Marist as one large poll, that would have 4,559 respondents (which results in a margin of error of 1.45%), with Ernst leading by roughly 1.3%. While that is still within the margin of error, and thus not enough to definitively say Ernst is ahead, the margin of error only refers to a 95% confidence interval. Thus, I do not think it is unfair to say that Ernst probably has a lead, unless there is reason to believe in a systematic bias in every Iowa poll of the past two weeks.

[SPC]

This wasn't included in RCP..... just saying

Because it was done for a liberal group, and was probably skewed a couple points or so in their favor.

This probably shouldn't be in the database, as it's not an independent poll.

Funnily enough, RCP won't include a PPP poll commissioned by the LCV, but will include a Magellan poll commissioned by the right wing "Keystone Report" website. They're hardly the paragon of objectivity you guys seem to think they are.

Obviously RCP has a Republican bias when it comes to which polls they enter; hence why I find it so amusing when you guys cite performances relative to the RCP Average as a basis for believing Democratic candidates will "overperform the polls"