February 1, 2012

2012 Florida Republican Presidential Primary Pollster Ratings

The following table lists the accuracy of pollsters in the January 31, 2012 Florida Republican Presidential primary using the measure of polling accuracy proposed by Martin, Traugott, and Kennedy.

An accuracy measure of 0.00 indicates that the odds ratio of the survey results matches the odds ratio of the actual results (the closer to 0.00, the better).

There are no differences in the accuracy measures for the Quinnipiac, ARG, Reuters/Ipsos, Miami Herald, NBC/Marist, Rasmussen, SUSA, War Room Logistics, Sunshine State News Service, PPP, Monmouth/SUSA, Suffolk, and Insider/Advantage polls at the 95% confidence interval.

* The accuracy measures for the We Ask America, CNN/Time, Dixie Strategies/First Coast News, and the Florida Chamber of Commerce polls fall outside the 95% confidence intervals. The larger sample sizes for the We Ask America poll (1,188 interviews) and the Dixie Strategies/First Coast News poll (2,567 interviews) did not increase the accuracy of those polls.

** Suffolk University also conducted surveys in three bellwether counties in Florida:

In Martin county, the survey result was Romney 52%, Gingrich 24%, and the actual result was Romney 52%, Gingrich 26%, for an odds ratio of 1.08 and an accuracy measure of 0.08.

In Sarasota county, the survey result was Romney 45%, Gingrich 21%, and the actual result was Romney 51%, Gingrich 27%, for an odds ratio of 1.13 and an accuracy measure of 0.13.

In Volusia county, the survey result was Romney 52%, Gingrich 24%, and the actual result was Romney 45%, Gingrich 36%, for an odds ratio of 1.73 and an accuracy measure of 0.55. The accuracy measure for this survey falls outside the 95% confidence interval.

Using the results from these three counties and the results from its statewide survey, Suffolk predicted that Romney would exceed 50% of the vote in the primary. Not only did the survey in Volusia county miss the actual vote, but also these three counties were not bellwethers for the 2012 primary. Seminole county was the closest in this primary, with Romney at 47% and Gingrich at 33%.

Details can be found here.

Averages-to-Date

The following table indicates the average absolute accuracy for polls in Iowa, New Hampshire, South Carolina, and Florida, and Iowa, South Carolina, and Florida.

January 31, 2012

Florida Republican Presidential Primary

Mitt Romney continues to lead the Florida Republican presidential primary. Romney leads with 43%, followed by Newt Gingrich with 31%, Rick Santorum with 13%, and Ron Paul with 9%.

In a similar survey conducted January 27-28, 2012, Romney was at 43% and Gingrich was at 32%.

A total of 36% of Republican primary voters say that they have already voted by absentee ballot or in early voting. Of this group of voters, Romney leads with 51%, followed by Gingrich with 29%, Santorum with 12%, and Paul with 8%. Among the 64% of the remaining likely Republican primary voters, Romney leads with 39%, followed by Gingrich with 32%, Santorum with 14%, and Paul with 10%.

Based on the votes already cast through absentee ballots and early voting and assuming Romney holds his vote, Gingrich would need to receive 51% of the votes cast on election day to tie Romney at 43%. That represents all of the undecided vote plus Ron Paul's expected vote on election day or 71% of Santorum's expected vote on election day.

Gingrich leads with 38% among likely Republican primary voters saying they are supporters of the Tea Party, followed by Romney with 35% and Santorum with 17%. Among likely primary voters saying they are not supporters of the Tea Party or are undecided about the Tea Party, Romney leads with 52%, followed by Gingrich with 23%, Paul with 15%, and Santorum with 8%.

Romney leads among men with 37%, followed by Gingrich with 33%, and Paul and Santorum with 12% each. Romney leads with 49% among women, followed by Gingrich with 26% and Santorum with 14%.

Details from the survey of 600 likely Republican primary voters conducted January 29-30, 2012 can be found here.

January 27, 2012

Arizona Republican Presidential Primary

Newt Gingrich and Mitt Romney are tied with 32% each in the Arizona Republican presidential primary. Gingrich and Romney are followed by Ron Paul with 12% and Rick Santorum with 10%.

Gingrich leads Romney 33% to 29% among likely Republican primary voters saying they will definitely vote in the February 28 primary. Romney leads Gingrich 56% to 24% among those saying they will probably vote.

Gingrich leads with 40% among likely Republican primary voters saying they are supporters of the Tea Party, followed by Romney with 23%, Santorum with 13%, and Paul with 11%. Among likely primary voters saying they are not supporters of the Tea Party or are undecided about the Tea Party, Romney leads with 40%, followed by Gingrich with 25%, Paul with 13%, and Santorum with 7%.

Gingrich leads Romney 35% to 30% among men, followed by Paul with 12% and Santorum with 9%. Romney leads Gingrich 34% to 28% among women, followed by Paul with 12% and Santorum with 11%.

Details from the survey of 600 likely Republican primary voters conducted January 25-26, 2012 can be found here.

January 23, 2012

2012 South Carolina Republican Presidential Primary Pollster Ratings

The following table lists the accuracy of pollsters in the January 21, 2012 South Carolina Republican Presidential primary using the measure of polling accuracy proposed by Martin, Traugott, and Kennedy.

An accuracy measure of 0.00 indicates that the odds ratio of the survey results matches the odds ratio of the actual results (the closer to 0.00, the better).

There are no differences in the accuracy measures for the ARG, PPP, Clemson, We Ask America, YouGov/Polimetrix, InsiderAdvantage, and the Rasmussen polls at the 95% confidence interval.

As noted below, however, the accuracy measures for the We Ask America, InsiderAdvantage, and Rasmussen polls fall outside of their respective 95% confidence intervals. The larger sample sizes for these polls (988 for We Ask America, 718 for InsiderAdvantage, and 750 for Rasmussen) did not increase their accuracy measures.

The accuracy measures for the ARG and PPP polls differ from the accuracy measures of the Politico/Tarrance, NBC/Marist, CNN/Time, Monmouth/SUSA, and the IPSOS/Reuters polls at the 95% confidence level.

* The accuracy measures for the We Ask America, InsiderAdvantage, Rasmussen, Politico/Tarrance, NBC/Marist, CNN/Time, Monmouth/SUSA, and the IPSOS/Reuters polls fall outside the 95% confidence intervals.

Details can be found here.

The following table indicates the average absolute accuracy for polls in Iowa, New Hampshire, and South Carolina and Iowa and South Carolina.

January 21, 2012

Obama Job Approvals and the National Economy

A total of 41% of Americans say they approve of the way Barack Obama is handling his job as president and 55% say they disapprove of the way Obama is handling his job according to the latest survey from the American Research Group. In December, 46% of Americans approved of the job Obama was doing and 51% disapproved.

When it comes to Obama's handling of the economy, 35% of Americans approve and 60% disapprove. In December, 42% approved and 54% disapproved.

Among Americans registered to vote, 42% approve of the way Obama is handling his job as president and 54% disapprove. On Obama's handling of the economy, 35% of registered voters approve and 60% disapprove.

Details from the nationwide survey conducted January 17-20 are available at The National Economy.

Why doesn't most advertising work? Take a look at 10 Rules for More Effective Advertising.

The margins of error reported for most polls underestimate the actual margins of error for ballot estimates from the same samples. Use the Ballot Lead Calculator to determine if the lead for any survey is statistically significant.