A recent online poll showed Republican Tom Foley leading Governor Dannel Malloy by nine percentage points in the race for the state's top office. The nature of the poll itself, though, is stirring discussion. Here's why.
The old-school poll worked like this. Pollsters would call you at home and ask you some questions about candidates. Fewer than ten percent of people who were called would participate. Eventually, cell phones were integrated. From those conversations, a series of conclusions would be made about politics, policy, and more. That's how places like Quinnipiac University do their polling.
Then there are polls like YouGov, the one run for The New York Times and CBS that showed Foley with the lead. It's done only on the Internet.
Scott Keeter, director of survey research for Pew Research Center, said there are two big differences between phone and Internet polling. First, online-only polling has a certain bias against the old and the poor -- people who don't have good access to the Internet.
"The bigger difference, though, between the YouGov polls and the kind of polls that Quinnipiac does or that Pew Research has traditionally done, is that the sample is what's called an opt-in sample," Keeter said. "People volunteer to be participants in YouGov."
That is, old-school polling chooses you. In new-school polling, you choose it.
"This is extremely controversial in the polling world right now," he said. "And the fact that The New York Times and CBS decided to switch, or at least to add to their polling, YouGov polls at this point is very big news."
Keeter said the YouGov poll is has a good track record, and is well-respected. He also said its recent Connecticut poll got a lot of attention, because it was the first one done since the spring.