Should you believe the polls?

[longknife]

My immediate response on seeing this was, “Hell no!” Read this”\

Polls show Trump climbing. But he still has a narrow path to win.” “Hillary Clinton Regains Momentum Against Donald Trump: Poll.” “Poll shows Clinton ahead of Trump in Florida, but race getting closer.” “Poll: Clinton regains 5-point lead nationally.” These are headlines that were all run by reputable news sources on the very same day, illustrating just how pervasive—and confusing—political polls can be in an election year. Polls dominate the news, inform forecasting models designed to predict the outcome of an election, and spur discussion and debate across the electorate.

How can anyone believe any of this garbage? And then we have those who live and die by their Electoral College predictions. As I've posted time and time again, the only thing to believe will be the results on the morning of November 9th.

This story @ Should you believe the polls? | Brookings Institution

 

[320 Years of History]

How can anyone believe any of this garbage?

By looking at the poll questions, demographics of the people polled, and poll methodology. That's what one needs to do, along with understanding mores and mechanics of statistical sampling (brief overview), prior to deciding one believes or disbelieves the results of a poll. (It does no good to look at that stuff if one doesn't know what are the implications of what one discovers when one reads it.)

It's important to look at that information. It's critical to knowing when to accept the results of a poll and when not to. Reputation of the pollster is not enough; even highly reputable pollsters have "moments" and make mistakes in the design and execution of their surveys. They don't necessarily do that often, but they do do it. One must check to find out whether they did it with each and every poll outcome pollsters publish.

 

[bill718]

My immediate response on seeing this was, “Hell no!” Read this”\

Polls show Trump climbing. But he still has a narrow path to win.” “Hillary Clinton Regains Momentum Against Donald Trump: Poll.” “Poll shows Clinton ahead of Trump in Florida, but race getting closer.” “Poll: Clinton regains 5-point lead nationally.” These are headlines that were all run by reputable news sources on the very same day, illustrating just how pervasive—and confusing—political polls can be in an election year. Polls dominate the news, inform forecasting models designed to predict the outcome of an election, and spur discussion and debate across the electorate.

How can anyone believe any of this garbage? And then we have those who live and die by their Electoral College predictions. As I've posted time and time again, the only thing to believe will be the results on the morning of November 9th.

This story @ Should you believe the polls? | Brookings Institution

Difficult to say. Polls are just a snapshot in time. 4 years ago, the polls predicted a big Romney win. Remember what happened?

 

[jwoodie]

Who pays for these polls? I refuse to participate in them, since their principal use is to affect election outcomes.