Apr 252010
 

Click here to read the Orange Juice Post on the Poll.

Adam Problosky is a pollster from Southern California who has been doing polls for several years.

The poll was touted by the Chuck DeVore for US Senate Campaign as a sign of strength for Chuck. It showed Campbell at 31, Carly Fiorina at 17 and Chuck DeVore at 14.

There is something called a sample when you do a poll.

When polling on a primary – you generally want to talk to people of the party in question. When polling – you typically ask likely voters.

OOPS – out of 751 Voters surveyed, only 272 Republicans.

All were newly registered voters in 2008/2009 – and are therefore not considered likely to vote in 2010.

Meg Whitman better not pay attention either…

Epic Fail.

  5 Responses to “Adam Problosky’s Poll… Epic Fail (Re: California US Senate Race)”

  1. Thanks Aaron, keep me informed. I have other campaigns I’m working on, so how much energy I needd to put into Logue’s matters.

  2. Hey Aaron, when I read your comment that “All were newly registered voters in 2008/2009 – and are therefore not considered likely to vote in 2010.” – I thought whoa, and downloaded a PDF of the poll from http://probolskyresearch.com/research.html – can you clarify what you meant?

    * Page 26: 304 persons participating in the GOP primary (272 Republicans, 32 Decline to State).
    * Page 28: shows all 304 as likely voters based on a ranking of 3-5 based on participation in the last five statewide elections (11/06, 2/08, 6/08, 11/08, 5/09)
    * Page 28: breakdown of those 304 voters by registration date is 20+ years = 61; 10-19 years = 79; 5-9 years = 99; 1-4 years = 56; <1 year = 9.

    Statistical research is an area where I have some expertise, therefore I always analyze sampling methods of polls. In this particular poll the sample size relative to the population is quite strong. This poll was intended to gauge the attitude of the California electorate, and polling 751 persons at a 95% confidence level gives a margin of error of only 3.7% (pg. 4). Anyway, not looking to jump on you, just trying to make sure a fairly conducted poll isn't unnecessarily sullied.

  3. Ken – was going based off of the criticism relayed to me.

    As an aside, If a sample of 272+32 is a valid sample – then wow.

  4. keep in mind the sample set isn’t 304, but 751 because the poll asked questions of likely voters of all political parties. As a comparison, a similar nationwide Gallup poll would have about 1300-1500 persons, with very large samples at about 4,000 to get to a +/-2% margin of error.

    What I found most disturbing about the poll is that Decline to State (DTS) voters are more inclined to request a Democratic over a a Republican ballot, even though the competitive races are on the GOP side of the aisle – that means we’re doing something wrong with marketing the party.

  5. Barnes – doesn’t it sound absurd to poll non-Republicans in a poll about a Republican Primary?

    Enough with beating a dead horse – but Adam’s poll was they only one out of 12 that had DeVore in double digits.

Sorry, the comment form is closed at this time.