Yahoo News Explains: How do pollsters get their numbers?

It all begins with a phone call.

According to the Pew Research Center, people are selected by “random digit dialing” of landlines and cell phones.

Typically, the sample size for a national survey is about 1,500 people. This small group is used to help determine how the majority of Americans feel about a certain topic.

Researchers have a 95 percent confidence level for most polls and believe they’re plus or minus three points of the true population opinion, but the center says one of the hardest parts of polling is getting people to actually talk to surveyors.

The percentage of people who participate has been steadily declining over the past decade.

There’s a one in 154,000 chance of being called, and volunteers aren’t used to ensure accurate, randomized results.

Internet polling can often be unreliable because not every demographic has access to it.

People with lower incomes, living in rural areas, or ages 65 or older are underrepresented. To combat this, researchers send surveys by other methods (like mail and telephone).

Most polls predicted Hillary Clinton would get more votes than Donald Trump — and technically, she did. However, polls were off from the final tally by about 2.7 points.

The Pew Research Center says this could be because:

  • Donald Trump’s supporters were less likely to answer pollsters.

  • Some Trump voters might not have revealed their support for him for fear of social scrutiny.

  • Models predicting who would vote could have been wrong, thus affecting who was polled.

The center’s director of survey research, Courtney Kennedy, believes polls are fundamentally accurate, despite the 2016 election cycle results. “The polling community and poll consumers should take some comfort in the fact that we’ve figured out quite a bit about what went wrong and why … and we all can learn from those errors.”

_____

Read more from Yahoo News: