A new polling technique on the horizon.

There is a good piece from the Pew Research Center discussing the decision by the New York Times and CBS News to engage a polling organization to create a panel for election coverage.

The New York Times and CBS News made big news in the polling world this weekend when they announced that they will begin using online survey panels from YouGov as part of their election coverage. YouGov, a U.K.-based research firm founded in 2000, uses such panels rather than traditional telephone surveys; the panel the Times and CBS are using has more than 100,000 members. The Times, citing concerns about the dearth of high-quality, non-partisan survey data, particularly at the state level, says it plans to include YouGov results as part of “a diverse suite of surveys employing diverse methodologies.”

One of the concerns is that the YouGov panel will use the Internet rather than traditional telephone surveys. The site indicates that 89% of Americans now use the Internet. (I wonder how many Americans don't have telephones, or have phones but refuse to answer calls from polling organizations.)

I have a concern that samples should be constructed according to what you want to learn. Thus, election polling often is more interested in learning what likely voters think than what people who are not registered voters think. A random sample of the population will include both groups. A sample of Internet users will tell you something about Internet users that would be harder and less accurate to infer from a telephone survey.

Would it not be nice to have a sample of opinion leaders polled in early 2015 about the November 2016 presidential elections? How would you construct such a sample.