National, statewide, and local polls always hold biases, especially in the face of an election. Regardless of the format or where the poll is taking place, there will always be undercoverage or people who don’t answer truthfully, which skews the results.
We talked to AP statistics teacher Alejandro Tyler and AP psychology teacher Jeff Corson about the data we collected. Since only about 10% of the student body responded to the survey, there inevitably is undercoverage and response bias in the results.
Tyler said that people with stronger political opinions are most likely to respond to the survey.
The low information voters, or really undecided voters, which is kind of the same group of people, they’re like, ‘I don’t know enough about politics to fill out their survey, so I’m not going to bother,’” he said. “Or ‘I don’t have any strong feelings about politics, I’m not going to fill out their survey.’”
Corson added that younger people under the age of 18 are usually impacted the most by their parents which impacts the data.
“And oftentimes when you’re younger, which is the results you’re getting, the people that can’t vote, they’re more influenced by their parents,” he said.
According to 538, leading up to the election 2016, it was predicted that Hillary Clinton would beat Donald Trump in the election 458 electoral votes to 80 based on national polls, which was obviously incorrect. Tyler said that due to bias, people were less likely to admit they were voting for Trump.
He said that in a case like this where there is so much response bias, it is fine to add a number onto that result in order for it to be more accurate.
“We should increase that number to represent the people who aren’t telling us the truth,” he said. “It was this percentage in 2016 so let’s add that on to our poll and say this is the real number.”
Your Vote Is Your Voice: https://arapahoextra.com/14376/uncategorized/your-vote-is-your-voice/
AHS Student On Why they are Voting: https://arapahoextra.com/14380/uncategorized/ahs-students-on-…-they-are-voting/
