Greensboro, NC -- With the political season in full swing, polls have been dominating coverage of the primaries.
Ever wonder how they come up with all those numbers? And why do different pollsters get different results on the same topics? Or why, perhaps, you or people you know never seem to get called about your opinion?
We took those questions to pollsters here in the Triad and found some interesting answers.
Think of polling as a snapshot of what people are thinking at a certain point in time. And here's it works:
Like the lottery, everyone has the same probability of being called for a poll.
Pollsters buy random lists of phone numbers, including landlines and cell phones. At the Elon University polling center, students make the calls.
On the other hand, Public Policy Polling uses interactive software. Next comes the data crunching stage.
And this is where we see different results. Why? Because there are so many variables that can influence your response:
"We try to construct a survey instrument that does not have bias," explained Mileah Kromer, Assistant Director of the Elon University Poll. "So we're really careful to avoid bias terms. For example, welfare is a word that certainly has a negative connotation to it. So often times we'll call things social assistance instead of welfare."
Kromer continues with an example that may come up this election season.
"A really good example of what could potentially bias, for example president Obama's numbers. Now, if you ask about the economy before asking about president Obama, you're priming the respondent to think about the economy and evaluate president Obama in terms of the economy. However, if you ask about foreign policy, before you ask about president Obama, you might get dramatically different numbers."
So, one variable is the wording. And in no particular order, second is cell phone versus landline.
Cell phone polls tend to generate responses from younger people who may have a different outlook than an older person.
Third, timing: doing a poll before or after a primary debate could give you different answers.
Fourth, of course the honesty of the respondents which greatly impacts accuracy?
And also, sample size: how many people did you survey and is the sample representative of the larger population?
"How can you tell me what's on North Carolinians minds by sample 500 or 800 people?" Kromer asked rhetorically. "It's the same kind of concept between how doctors can tell what your blood type is from a random sample of blood. So if the sample is drawn randomly, I can tell what's on Carolinian's minds."
To cap it off, there's no "Do Not Call" list for polling. That's because public opinion polling is not considered telemarketing.In short, if you have a registered number, chances are, you will be getting a call sometime.
The polls are sure to continue through primary season which picks up next week.
After the Florida primary there will be the Nevada and Maine caucuses on February 4th. Then, just three days later, people in Colorado and Minnesota will vote in their caucuses while Missouri holds its primary.
Followed by the caucus in the Northern Mariana Islands on the 25th and the Arizona and Michigan primaries on the 28th.
WFMY News 2