Re: TIMES/BLOOMBERG POLL,Obama holds 12-point lead over McCain, poll finds

On Jun 24, 5:21 pm, Jim Higgins <gordian...@xxxxxxxxxxx> wrote:
allan.san...@xxxxxxxxx wrote:
On Jun 24, 4:16 pm, Rita <R...@xxxxxxxxxxx> wrote:
On Tue, 24 Jun 2008 18:11:58 -0400, Jim Higgins

<gordian...@xxxxxxxxxxx> wrote:
TIMES/BLOOMBERG POL Obama holds 12-point lead over McCain, poll finds
http://www.latimes.com/news/politics/la-na-poll25-2008jun25,0,5763707....
2:00 PM PDT, June 24, 2008
WASHINGTON -- — Buoyed by enthusiasm among Democrats and public concern
over the economy, Sen. Barack Obama (D-Ill.) has captured a sizable lead
over Sen. John McCain (R-Ariz.) at the opening of the general election
campaign for president, the Los Angeles Times/Bloomberg Poll has found.
In a two-man race between the major party candidates, registered voters
chose Obama over McCain by 49% to 37% in the national poll conducted
last weekend.
I wonder why this poll and one other -- Newsweek -- show Obama with a
hefty lead and Gallup daily polling keeps finding only a 3 point or
so difference? Not that I'm complaining if it is actually so:)

Usually these differences are attributable to sampling frame.
Typically this bias is accommodated by the statistical margin of error
(when they mention +/- X%). In this case, the frame error or the
sampling error were reported but the distribution of respondents was.

"In this national poll's random sample of voters, 39% identified
themselves as Democrats, 22% as Republicans, and 27% as independents."

This is not to imply that the poll is wrong because it included more
Democrats than Republicans - please don't misunderstand - though in
the absence of knowing the population distribution of D/R/I, you just
don't know the direction or magnitude of any bias that this may have
caused. If in fact it was a random sample one would hope that there
was no systematic bias.

al

Another factor is when the poll(s) were taken. McBush sure seems to be
slipping but it is a long time until November so McBush can slip even more.

--
Civis Romanus Sum

Good point. Most of these types of polls are conducted over a two or
three day period and involve somewhere in the 900 to 1200
respondents. The samples are typically random digit and tend to
oversample higher socio-economic status - it is rare to get a homeless
person answering a phone. But they make great sound bites!

al
.

Relevant Pages

  • Polling Memo: The Shift to Hillary
    ... Barack Obama recently declared himself the frontrunner in the race, ... Democrats by 7 points, and the latest Zogby/Reuters poll has Sen. ... Barack Obama doing better than Hillary against Sen. McCain. ...
    (rec.outdoors.rv-travel)
  • Re: TIMES/BLOOMBERG POLL,Obama holds 12-point lead over McCain, poll finds
    ... Buoyed by enthusiasm among Democrats and public concern ... Sen. Barack Obama has captured a sizable lead ... the Los Angeles Times/Bloomberg Poll has found. ... If in fact it was a random sample one would hope that there ...
    (soc.retirement)
  • Re: Is the Newsweek Poll Biased Media?
    ... they polled Democrats 2-1 to achieve the 15 point lead they reported. ... That would would mean the poll is inaccurate and misleading. ... certainly distinguish bias as long as there is raw data available. ... there is no liberal bias in the media. ...
    (rec.music.gdead)
  • Re: Is the Newsweek Poll Biased Media?
    ... they polled Democrats 2-1 to achieve the 15 point lead they reported. ... That would would mean the poll is inaccurate and misleading. ... certainly distinguish bias as long as there is raw data available. ... there is no liberal bias in the media. ...
    (rec.music.gdead)
  • Re: TIMES/BLOOMBERG POLL,Obama holds 12-point lead over McCain, poll finds
    ... Buoyed by enthusiasm among Democrats and public concern ... Sen. Barack Obama has captured a sizable lead ... Typically this bias is accommodated by the statistical margin of error ... They also don't reach cell phones. ...
    (soc.retirement)