Poll games

Rosslyn Smith
Both Big Government and NROs Campaign Spot have noticed what seems to be a trend in recently released polls on political issues and possible 2012 presidential matchups: The use of random samples of "adults" as opposed to "registered voter" or "likely voters".  As Dr, Susan Berry at Big Government notes:

Keep in mind that likely voters are not representative of the general population of telephone owners. Likely voters are generally more aware and informed about the issues, enough to commit to get out and vote.

I anticipate any number of games will be played with media sponsored polls in the coming months to make Obama's numbers look better.  In addition to the composition of the sample, the wording of questions, the order in which they get asked and the omission of a key question can all skew the results.  More than ever it pays to go to the poll itself and examine the details instead of relying on the media's summary.

 

Questions on policy details are particularly susceptible to bias in the order and wording of questions. Consider a recent Quinnipiac poll of registered voters on the debt ceiling impasse.  The media has used this poll to show that voters would blame Republicans more than they would Democrats for a failure to raise the debt ceiling.  The first question on economic priorities offered a false choice.

 

What do you think is more important - reducing the federal budget deficit, or reducing unemployment? 

Quinnipac then prefaced the question on which political party should bear the blame if the debt ceiling is not raised with this statement.

Unless Congress and the president can agree to raise the federal debt limit soon, the government will not be able to borrow more money to fund its operations and pay its debts.

I suspect the answers would be different if the question had been worded "Should Congress raise the debt ceiling so the federal government can continue to spend $1 for every 58 cents it takes in."  And of course, Quinnipac's question on future tax increases specified there were to be borne by corporations and the wealthy.  

Both Big Government and NROs Campaign Spot have noticed what seems to be a trend in recently released polls on political issues and possible 2012 presidential matchups: The use of random samples of "adults" as opposed to "registered voter" or "likely voters".  As Dr, Susan Berry at Big Government notes:

Keep in mind that likely voters are not representative of the general population of telephone owners. Likely voters are generally more aware and informed about the issues, enough to commit to get out and vote.

I anticipate any number of games will be played with media sponsored polls in the coming months to make Obama's numbers look better.  In addition to the composition of the sample, the wording of questions, the order in which they get asked and the omission of a key question can all skew the results.  More than ever it pays to go to the poll itself and examine the details instead of relying on the media's summary.

 

Questions on policy details are particularly susceptible to bias in the order and wording of questions. Consider a recent Quinnipiac poll of registered voters on the debt ceiling impasse.  The media has used this poll to show that voters would blame Republicans more than they would Democrats for a failure to raise the debt ceiling.  The first question on economic priorities offered a false choice.

 

What do you think is more important - reducing the federal budget deficit, or reducing unemployment? 

Quinnipac then prefaced the question on which political party should bear the blame if the debt ceiling is not raised with this statement.

Unless Congress and the president can agree to raise the federal debt limit soon, the government will not be able to borrow more money to fund its operations and pay its debts.

I suspect the answers would be different if the question had been worded "Should Congress raise the debt ceiling so the federal government can continue to spend $1 for every 58 cents it takes in."  And of course, Quinnipac's question on future tax increases specified there were to be borne by corporations and the wealthy.