NYT Polling Bias

The New York Times recently released a poll showing Obama’s popularity spread much higher than any other active poll at +37.  With a margin of error of only +/-3% how can this be explained?

In the past, I have discussed how “Likely Voter” polls give more accurate results than “Registered Voter” polls or “Adults.”  The Rasmussen Likely Voters poll currently has Obama at only +10% with their Presidential Approval Index at only +2%.  The New York Times uses “Adults.”

Even among “Adults” polling, the New York Times poll is still way outside the norm.  Within 24 hours of the NYT/CBS poll a NBC/WSJ “Adults” poll only had Obama at +22%.  The key is with the statistical sampling they don’t print in their “news” article.

When asked who they voted for in 2008 this was the NYT sample:

Obama: 48%

McCain: 25%

Other: 1%

Refused: 5%

Didn’t Vote: 19%

Obama won the actual popular vote by a 52.87%-45.61% margin.  Not by a 2:1 margin.  The actual results of the election are not subject to opinion or any margin of error.  They are a known quantity.  With the known numbers right, the NYT numbers must be wrong.

MORE BELOW THE FOLD….

Even if we eliminate those who did not vote or refused, the actual voting replies calculate as follows:

Obama: 64.86%

McCain: 33.78%

Other: 1.35%

The NYT sample shows an Obama victory 4 times the size of his actual victory.  McCain fared better than these national numbers even here in Massachusetts.

The NYT poll shows only 26% of the people voting against Obama also shows his negatives at 26%.

In reality over 47% of the people voted against Obama.  What are his actual negatives?

(BTW:  The NBC/WSJ also over polled Obama voters, but by a much more reasonable margin)

About Mike "DD4RP" Rossettie