As the end of voting draws near next month, a number of new polls are showing where the election results might be heading.
Putting together a poll involves talking to a bunch of people and then figuring out how to draw your sample.
"It's a little bit science, it’s a little bit art," says Rebecca Bromley-Trujillo of Christopher Newport University.
"And so, you want to start by drawing your sample among folks that have been consistent voters. So, for us, they need to have voted at least twice in the last four general elections. And then we also do pull in a subsample of newly registered voters. Not all pollsters do that, but we do want to make sure we're capturing some of the new people that are entering the voting market, if you will."
Statewide polling is usually more accurate because it's easier to get a large sample size. Chaz Nuttycombe runs polling for State Navigate, and he says the trick is to balance all the variables.
"You could talk about margin of error, you could talk about sampling size, you could talk about how it's weighted. Is it weighted by region? That's something we do at State Navigate," Nuttycombe says. "We try and make sure that we weight a poll by either county or a group of precincts just to make sure that we have the appropriate amount of rural voters and suburban voters and urban voters."
One of the trickiest decisions is how many Democrats and Republicans to include, a variable that can change with how motivated people might be to cast a ballot.
