A recent poll has sparked some debate in Oregon conservative circles on social media. Unfortunately, polling is still widely misunderstood by a lot of people. That’s why I wrote a two
series about polling back in 2015.
One thing I didn’t cover in that series is polling methods. I’ll do my best to cover that topic now.
It’s important to remind everyone that true internal polls are never released to the public. Occasionally portions of internal poll numbers are used in a polling press release, but I consider these to be a different category of polls (I call them persuasive polls).
When a campaign releases a persuasive poll, it’s because they intend to persuade voters and the press to believe the narrative they’ve constructed. That doesn’t mean the poll is wrong, but it often means the premise of the poll isn’t as objective as it could be. Persuasive polls can still be wrong and often provides an incomplete narrative about the state of any particular election.
One of the most important things you must consider when looking at any poll is the manner in which the poll was conducted. There are many ways to conduct a poll. Live calls, automated calls, and Internet surveys are the most common.
Internet surveys are unproven but are slowly becoming more reliable for certain kinds of national polling. I am still skeptical of their accuracy in statewide or smaller contents.
Live telephone polls are the most accurate form of polling but are more expensive to conduct. They are currently suffering from lower response rates (which is a major factor that can determine the cost and accuracy of poll).
Automated polling is less expensive but less accurate for a few reasons. For instance, it is illegal for an automated poll to call cell phones. This could mean the results are skewed (particularly in Democratic primary races) because there are no respondents from cell phone-only voters. Automated polling also can’t verify that the target voter is the one actually answering the poll questions which results in bad data and more unreliable results.
There isn’t a lot of polling in Oregon, so even bad polls are apt to generate some buzz. In Oregon’s polling desert, it’s essential that we don’t drink immediately from any source of water we see. Some of them will be perfectly fine to drink, but others can make you sick or crazy.