***Updated*** AP-Yahoo poll is biased in seven “distinctly invalid ways relative to known demographics”
I'm including all the relevant poll information in this diary entry.
In DJ Drummond’s blog discussing the latest AP-Yahoo poll, he stated that the poll was “biased in six distinctly invalid ways relative to known demographics”. So I decided to look for myself and I found 7 worth noting. And I compared his “known demographics with census information. These are the results from Knowledge Networks who conducted the poll, matched with DJ Drummond’s analysis. I made the type bold for DJ Drummond to make it easier to read.
First here are my sources:
A blog by DJ Drummond http://tinyurl.com/5jm2en (thanks PaRep),
A link to yahoo http://news.yahoo.com/page/election-2008-political-pulse-souring-on-mccain just click where it says “Full poll results”
UPDATE I was unable to produce the full tables due to formating, so I limited the sample to the first wave of calls. There is very little change from Wave 1 to Wave 8
Census info on 2004 election http://www.census.gov/prod/2006pubs/p20-556.pdf You have to scroll down to Table B.
Here are the seven fundemental problems with this poll, starting with the least objectionable.
*1. DJ Drummond – LV5 asks if the person is registered to vote. 86% say yes, so that tells you that unless the question was only asked of registered voters, one of every seven answers is irrelevant. *
Knowledge Networks – LV1. Are you registered to vote?
Yes, at my current address 76%
Yes, but at a different address 7%
Not sure 3%
Census.gov – Registration and Voting 18 and older
Registered Voters (142,070,000)
Actual Voters (125,736,000) Note: 88% of those registered, voted.
Pop Eligable to Vote (215,694,000)
2. DJ Drummond – 22% of respondents are identified as 18-29 in age, versus 17% in that age group’s actual voting in 2004.
Knowledge Networks – AGE
Census.gov – Age
18-24, 9% (11,639,000)
25-34, 14% (18,285,000)
Total Voted (125,736,000) Note: All %s will be calculated using Total # Voted, with Income the only exception.
3. DJ Drummond – In this poll, 13% of the respondents did not complete High School, versus only 4% of voters in 2004 who did not complete High School. 31% of the poll respondents have a High School diploma as their highest education, versus 22% of voters in 2004 in that category. The overweight is obvious.
Knowledge Networks – EDUCATION
Less than high school 13%
High school 32%
Some college 28%
Bachelors degree or higher 27%
Census.gov – Educational Attainment
4. DJ Drummond – Whites in the poll made up 69% of the pool, versus 77% of voters in 2004. Hispanics in the pool counted for 13%, versus 7% who voted in 2004.
Knowledge Networks – RACE / ETHNICITY
White, Non-Hispanic 70%
Black, Non-Hispanic 11%
Other, Non-Hispanic 5%
2+ Races, Non-Hispanic 1%
Census.gov – Race and Hispanic Origin
White alone, non-Hispanic, 79% (99,567,000)
Hispanic (any race), 6% (7,587,000)
5. DJ Drummond – In the poll, 84% of respondents live in urban areas, versus 30% of actual voters in 2004.
KN – METROPOLITAN STATUS
Census.gov – not available in this report.
6. DJ Drummond – 58% of respondents to the poll make $50,000 a year or less, versus 45% of the actual voters in 2004.
Knowledge Networks – INCOME
Under $15,000 15%
$15,000 to less than $20,000 5%
$20,000 to less than $25,000 7%
$25,000 to less than $30,000 6%
$30,000 to less than $40,000 15%
$40,000 to less than $50,000 12%
$50,000 to less than $75,000 21%
$75,000 to less than $100,000 11%
$100,000 or more 8%
Census.gov – Annual Family Income of Actual Voters
<50,000, 33% (32,852,000)
Total family members, (97,352,000)
And my personal favorite…
**7. DJ Drummond – PID1 notes that the party affiliation split for the poll was 40% Democrat, 27% Republican, 21% Independent, and 12% undecided or other, which actually creates a DRI split of 40/27/33, as usual for this year far out of balance with historical norms.
PARTYID goes further, noting that 49% of the respondents consider themselves Democrats, 37% Republicans, and just 14% other or Independent, a contrast to PID1 which is not explained. **
Knowledge Networks – REG1. Do the state records show that you are currently registered as a Republican, a Democrat, an
Independent, or do voters not register by party in your state?
I’m registered as a Democrat 36%
I’m registered as a Republican 29%
I’m registered as unaffiliated or
My state does not allow registration
with a party 4%
I’m not sure 16%
Knowledge Networks – PARTYID. Do you consider yourself a Democrat, Republican, an Independent, a supporter of some other party, or none of these?
Some other party 4%
None of these 8%
Census.gov – not available in this report, but of course it’s been the subject of many blogs and news articles. I’m pretty sure it’s been historically less than 12% Democrat advantage.
Its obvious the pollsters/Democrats are hoping for a different voting demographic than 2004.
If numbers exist on the demographics of where 2004 voters lived (ie. metro/non-metro), I will try to find them.
It would explain a lot if the pollsters are primarily polling in cities.
Updated Sorry about any formating errors, I’m relatively new to this type of format.