Calblog

KABC is a radio station. Maybe its just a poll of talk radio listeners or callers. If that's the case, I'm not surprised by the result.

This was KABC-TV poll (channel 7 here in LA). It was presented as the same random sampling poll as the others.

The devil is in the details, especially with an unpredictable election like this one. How were “likely voters” determined? Did the pollster ask any follow-up questions to try and convince undecided voters to settle one a candidate? What were the follow up questions? Over what period of time was the poll conducted? Who were the potential gubernatorial candidates named by the pollster? Was "Other" or "none of the above/ skip the question" an option?

Since about the only thing the LA Times poll was good for was its discovery that marginally undecided voters appear to lean towards Bustamante, it would be helpful to know how many undecideds are this poll. If the undecided number is around 50%, the poll probably can’t tell us who is gaining and who is fading. If the undecided numbers are more in line with the Field Poll (around 30%) then this poll would tend to indicate Arnold is pulling ahead of Bustamante.

Brother Sean, you have got to be doing this stuff for a living!

:-)

[sarcasm]
The Los Angeles Times is a left-wing newspaper. Maybe last week's poll was just a poll of people who read the L.A. Times. If that's the case, I'm not surprised by the result.
[/sarcasm]

Here's a link:
http://abclocal.go.com/kabc/news/082703_nw_ewn_poll.html

Speaking of polls whatever happened to Weintraub's looking into the 125 "phantom latino" likely voters in the LA Times poll. The slight explanation Wientraub of Cal Insider offered about this was confusing and admittedly just speculation. He said that the Internet was buzzing about this and that he had a call into the paper. Does anyone have any explanatory links or know if the LA Times responded? I've looked but haven't found a thing.

Much as I would like it to be accurate, I'm not sure that this poll is any more reliable than the LA Times Poll.

First of all, note the low number of respondents -- only 591. Most polls taken in the last few years have at least 1000 respondents. Also, look at the number of undecided voters -- only 4%! Think about this -- 6 weeks after an election is announced, three weeks after the candidates are announced and only 4% of the population doesn't know who they'll vote for? The number of undecided voters in last years election was higher than this the week before the election -- and that campaign lasted over 6 months. This makes the 15% undecided in the LA Times poll (which was itself only 1/2 the size of other polls) look puny.

And look at the way votes are counted -- all registered voters who say they plan to vote. Trouble is, people lie about performing their civic duty all the time (for example, a whopping 85% of people surveyed in a post-election 1988 poll claimed to have voted that year). Oddly enough, people are less likely to lie about whether they will vote in the future than whether they voted in the past, but simply taking voters word on their intentions isn't good enough for an accurate poll.

While I'm on the subject, here's some food for thought:

A lot of people who comment on this blog (Justene, xlrq, spooky, myself) appear to prefer McClintock, but are ready to support Arnold if he can prove himself and avoids collapsing like Simon did last year.

Here's my question to everybody for whom the above paragraph applies: If a pollster called YOU and asked YOU for your preference, what would you say? Tom? Arnold? Or undecided? And if (like the ABC and LAT polls) the pollster tries to push you out of the undecided camp, where would you go? For the record, I'd say Arnold, simply to keep his poll numbers high (since it could help create a self-fufilling prophecy), but what about the rest of you?

Speaking hypothetically, as I must, I would tell them my first preference. Otherwise you stand a very good chance of outsmarting yourself.

Schwarzenegger.

what would you say? Tom? Arnold? Or undecided?

Tom Arnold?

The last Field Poll had 629 registered voters, 448 of whom were deemed to be likely voters, but no one quibbled with its sample size. While I agree that this poll may not be any more representative of truth than the LA Times poll, that is because both are likely flawed.

and why is it that the only poll that explicitly makes reference to adding in a very large proportion of latino likely voters is also the only on that has Bustamonte in a large lead?

I'd say McClintock, even though the odds are better than 50-50 I'll end up voting for Schwarzenegger. Watering down your vote at the last minute based on the polls is one thing, and helping water down the polls themselves is another.

You can also check the pages about insurance quotes insurance quotes http://insurance-quotes.rulo.biz/ poker poker http://poker.rulo.biz/ debt consolidation loans debt consolidation loans http://debt-consolidation-loans.rulo.biz/ bad credit auto loans bad credit auto loans http://bad-credit-auto-loans.rulo.biz/ countrywide home loans countrywide home loans http://countrywide-home-loans.rulo.biz/ unsecured loans unsecured loans http://unsecured-loans.rulo.biz/ homeowners insurance homeowners insurance http://homeowners-insurance.rulo.biz/ canadian pharmacy canadian pharmacy http://canadian-pharmacy.rulo.biz/ online pharmacies online pharmacies http://online-pharmacies.rulo.biz/ canadian pharmacies canadian pharmacies http://canadian-pharmacies.rulo.biz/ reductil reductil http://reductil.rulo.biz/ butalbital butalbital http://butalbital.rulo.biz/ propecia propecia http://propecia.rulo.biz/ clonazepam clonazepam http://clonazepam.rulo.biz/ fluoxetine fluoxetine http://www.rulo.biz/ wellbutrin wellbutrin http://wellbutrin.rulo.biz/ slots slots http://slots.rulo.biz/ black jack black jack http://black-jack.rulo.biz/ craps craps http://craps.rulo.biz/ roulette roulette http://roulette.rulo.biz/ - Tons of interesdting stuff!!!