At ABC News.com, survey expert and Stanford professor Jon Krosnick has more on the likely primacy ballot effect that I reported on this morning:
Until this year, New Hampshire rotated candidate name order from precinct to precinct, which allowed us to do that analysis.This year, the secretary of state changed the procedure so the names were alphabetical starting with a randomly selected letter, in all precincts.
The randomly selected letter this year was Z.
As a result, Joe Biden was first on every ballot, Hillary Clinton was near the top of the list (and the first serious contender listed) and Barack Obama was close to last of the 21 candidates listed.
Thus, I'll bet that Clinton got at least 3 percent more votes than Obama simply because she was listed close to the top.
Most, if not all, of the pre-election telephone polls rotated name order from respondent to respondent, which meant name order did not distort their overall results. Failing to incorporate the name order effect that probably happened in the voting booth is therefore probably partly responsible for the polls' inaccuracy.
More importantly, if New Hampshire had rotated name order in the voting booth as it has always done in the past, the race would probably have been too close to call without a recount and might even have been an Obama victory.
Education & Careers
Politics
Comments
This may be so, but I think it doesnt completely explain how most polls underestimated clinton's support by about 10%, not 3%...though they got Obama's vote count dead on..
Most likely, it was a combination of factors such as
1. Primacy Effect
2. Women feeling sympathy for Clinton, especially after her debate and public crying episode.
3. Negative racial attitudes among some voters toward Obama
4. Rebelling against Iowa preferences
5. Greater name recognition and heuristic voting among low-income/low-educated voters.
Posted by: erik | January 9, 2008 1:09 PM
Is "Thus, I'll bet" a new statistical analysis program?
I do hope so, I'd hate to think you might be thinking of using some hand waving speculation to try to cover up the fact that you called this one completely wrong.
Posted by: Sigmund | January 9, 2008 1:28 PM
This year, the secretary of state changed the procedure so the names were alphabetical starting with a randomly selected letter, in all precincts. The randomly selected letter this year was Z.
So basically, given the lack of Z names, it was just normal alphabetical order?
More than that, this doesn't seem like proper randomization at all, alphabetically Examplename would be placed above Furthername in nearly all iterations, about 96%. And Examplename is always going to be placed above Extremename?
Sorry if I've missed some obvious nuance here.
Posted by: This is Common Sense | January 9, 2008 3:22 PM
This is all pointless drivel. Even if you flip the outcome, it's really the SAME OUTCOME.
Obama picked up 9 delegates.
Clinton picked up 9 delegates.
You need over 2,000 to win the nomination. No matter what you say, how you slice it - NH is ultimately close to worthless. Don't let the media hyperinflate states the size of suburbs.
Posted by: cephyn | January 9, 2008 4:10 PM
common sense: this doesn't seem like proper randomization at all...
Very true. And it just gets worse. Some letters are extremely uncommon for last names. Whoever has a name that comes just after a stretch of uncommon letters gets a higher chance of being first.
If they want to save money or something (presumably why they didn't print rotating ballots as previously?), why not just do true random shuffling? Why the need to stick to alphabetical?
Posted by: kevin | January 9, 2008 4:31 PM
Andrew Gelman has something on the polls too. I guess he'll discuss it more in the near future too. Worth reading, as he works on the analysis of polls and voting patterns.
Bob
Posted by: Bob O'H | January 10, 2008 10:04 AM
New Hampshire and Iowa are important, because they are first. And the natural instinct of humans is to side with the popular one. Whoever wins in those two gets a considerable advantage in the last month of campaigning before Super Tuesday. Whoever performs really poorly will see their supporters running off in search of a candidate to support who seems more likely to win.
Posted by: Suricou Raven | January 10, 2008 10:54 AM
More importantly, if New Hampshire had rotated name order in the voting booth as it has always done in the past, the race would probably have been too close to call without a recount and might even have been an Obama victory.
Thankfully, it won't be long before we see who actually gets it, regardless of these minor factors, ( that of course can make a big difference)!
Dave Briggs :~)
Posted by: Dave Briggs | January 10, 2008 12:41 PM