Journalists across the nation have been doing a lot of soul-searching since the 2016 election, concerned that poor coverage choices may have influenced the outcome.

"Nearly every major media outlet had spectacularly guessed wrong on the outcome, had failed to see the rise of the electorate that would elect Trump, and had not covered Trump as a serious candidate. Those sins, along with an obsession with Hillary Clinton’s emails and a wrongful dismissal of Russian involvement in the American election, pumped oxygen into the toxic political environment that helped produce the Trump presidency," write the editors of the Columbia Journalism Review. "This time around, the press has pledged to do better. Yet, with 11 months to go before Americans go to the polls again, there already are signs that journalists will repeat the mistakes of 2016."

So, CJR and The Guardian US teamed up to talk to 30 journalists who cover elections or monitor election coverage. Here's a sample what they had to say about 2016 coverage and what changes they plan or hope to see in 2020 on a range of subjects.

Sarah Kendzior, author of The View from Flyover Country: Dispatches from the Forgotten America, said coastal media outlets are "parachuting in here with the narrative pre-written trying to find people who fit their preconceptions of what people in the Midwest are like. . . . The best way they could fix this problem would be to hire people who actually live in these states."

The next quote comes from Al Cross, director of the University of Kentucky's Institute for Rural Journalism and Community Issues, which publishes The Rural Blog: "You can’t have a parachute mentality. You have to have some rural sensibility or an appreciation of rural sensibility. You deal with them as people and you have an appreciation for how they live their lives, and be respectful of that. And if you show that to them, they will show that to you."

Todd Gitlin, professor of journalism and sociology at Columbia University, said: "News media went from not noticing people whose life chances are impacted by the rustification of the Midwest to thinking that they are now the central story. It’s an absurd overcompensation. This is another instance of pack journalism, from ignoring a population to doubling down on a population. And in both cases, it’s a stampede reaction, not grounded in sociological knowledge or political science knowledge, but rather in a kind of feeling of having failed to get the odds right during the campaign."

In a discussion about polling, USA Today Washington Bureau chief Susan Page says, "You have to rely both on quantitative and qualitative interviews. Polling is really valuable, but it doesn’t tell you everything. You don’t want to go to a town, go to a diner, interview six people, and pretend like you know what people are saying. You want some quantitative data that enables you to say, ‘Hey, Pete Buttigieg is doing really well among older voters,’ then go out and talk to older voters about why that might be. In the USA Today/Suffolk poll, which is our big national poll, Joe Biden continues to do very well in the Democratic horse race. But we always do call-backs to people we polled to talk to them, to gather quotes. There’s just a different quality to the information you get in a conversation as compared to the data you get in a poll. And one thing we’ve noticed is that people who tell us they support Biden do not have the fervor of people who are supporting other candidates. And so by doing the callbacks, it makes us realize that that number is soft—we don’t know that it is going to change, but it might change."