The Gallup Tracking Poll That Is Freaking Some People Out Was A Three Day Tracker

[DemocratSinceBirth]

The tracker caught Romney at the height of his bounce, the absolute height of his bounce:

http://www.gallup.com/poll/157907/romney-narrows-vote-gap-historic-debate-win.aspx


If you followed elections you would see even wilder swings in three day trackers.

[7worldtrade]

1. Its actually a 7 day tracker - Rasmussen's 3 day tracker shows the bounce has stalled

Gallup uses a 7 day rolling average so any big swing in any direction will be muted at first. Gallup merely said that the the last 3 days of interviews have showed a 47-47 tie. the prior 4 days of their tracker had obama at 50 and Romney at 45, so Romney clearly had a bounce and Obama clearly tracked down. the only question is whether the numbers will move back in Obama's favor. If they do, Obama will stay up. If they stay a tie, then once the full 7 days are accounted for the topline numbers will be even.

Rasmussen's 3 day rolling average had Romney going from minus 2 to plus 2 over Obama between Thursday through Saturday. With Sunday's numbers it has gone to 48-48, a tie. This means that Romney had 2 great polling days Thursday and Friday and two less great polling days Saturday and Sunday. The only question is whether the trend will continue back to Obama plus 2.

http://www.rasmussenreports.com/public_content/politics/obama_administration/daily_presidential_tracking_poll

[DemocratSinceBirth]

2. Thanks For The Explanation

Does that mean there really wasn't a five point bounce?

Wouldn't you need two three day trackers, one before and one after the debate, to conclude that?

[My Pet Goat]

3. nope gallup broke it out 3 days

during wich it was tied meaning obama probably recovered somewhat on the weenkend.

[DemocratSinceBirth]

4. Now I'm More Confused

I would think the damage by the debates was mitigated by the repair from the good jobs numbers.

[My Pet Goat]

5. thats what i think happened, but

it is possible gallup measured approximately 47%-47% each of the three days.

[7worldtrade]

7. You have to pay attention to the averaging

Gallup's top line numbers that show Obama 49 and Romney 46 as of yesterday are based on 7 days of polling. Each day of polls accounts for 1/7 of the top line number. So the top line number as of yesterday reflects 4 days of pre-debate polling where Obama is ahead by 5 and 3 days of post-debate polling where they are polling even at 47-47. Unfortunately, according to Gallup, Obama and Romney were polling even at 47-47 on each day following the debate and the numbers didn't get better on Saturday. We don't know yet about Sunday - those numbers won't be reflected in the average until 1 pm today. But even if Obama starts gaining some its likely that the top number will still go down unless Obama was up by 5 on Sunday (that's because the new average will drop off one pre-debate day where Obama was up by 5 and add yesterday's number, which it would appear will not have Obama up by 5 if they were 47-47 even on Saturday, but I obviously don't know).

[DemocratSinceBirth]

13. If Nate Silver And Sam Wang Can't "Reverse Engineer" Tracking Polls Neither Can We

.

[TroyD]

6. RCP jumped with excitement to post the 47-47 Tie

On their website this morning.

But I'm assuming they're going to have to replace it at 1:00 PM, because that's when the proper daily Tracking Poll is released.

Yesterday Obama was at (+3), and if Romney came down today in Rasmussen from (+2) to a TIE, I'm assuming Romney isn't going to go up in Gallup.

[Floyd_Gondolli]

14. Surprised they got around to it

It means they had to take a break from aggregating essays from the Heritage Foundation and content from World Net Daily.

[hrmjustin]

8. I thought that Gallup reports their findings at 1 pm every day. Have they moved that up now...

... to the morning.

[TroyD]

9. No, their regular Tracking Poll comes out at 1:00 PM

If you go to the main Gallup page on the right side, it still shows the latest tracking poll from yesterday as Obama +3:

Obama - 49

Romney - 46

Please also read my post above where I explained it.

http://www.gallup.com/home.aspx

[hrmjustin]

11. yes it was strange when i saw it this morning.

[Jennicut]

10. It is also the 4th through the 6th.

Did they do new polling Sunday?

Sigh. I hate when polls try to influence things with half truths. Obviously Romney's bump has most likely stalled. Even in those few days he pulled even with President Obama he could not get ahead. He pulled even with Mittens on Rasmussen. Why not just wait until 1:00 to post the 7 day tracker with the new polling data? Unless they want to be a 3-4 day tracker.

[DemocratSinceBirth]

12. That's Why Objective Social Scientists Don't Rush Into The Field After A Major Event

Even if their findings are statistically sound they have been heavily influence by current events.


On 9/12/01 Bush had a 90% approval rate.