We spent much of 2016 and 2017 fretting about the effects of fake news and other forms of viral misinformation. Some did so by proclaiming the advent of a “post-truth” era.

So many of the pixels we dedicated to this topic, however, were uninformed or ill-informed. Not because of any intention to mislead on the part of the authors (usually), but because we are still woefully ignorant of the real effects of fake news and how to fix it.

Sure, we know anecdotally that misinformation can have real-world effects. “Fake news” led an armed man to “investigate” a pizzeria in D.C., a protester to heckle Emmanuel Macron outside a Whirlpool factory in Amiens, and blasphemy charges to be levelled against an innocent candidate in Jakarta. And we’ve long known that sometimes, just sometimes, politicians will retract or drop a debunked claim.

But there is still so little research that we can apply to everyday fact-checking work. We know quite a lot about the reach of fake news, but not much about its capacity to sway votes or affect decisions. We know at last that fact-checking probably doesn’t backfire in experimental settings, but not whether that is true in real life.

2017 saw a lot of new research in this space, to the point where the International Fact-Checking Network (IFCN) is launching a research database to catalog the most interesting academic findings. But it’s still not enough to help practitioners do a better job.

I am hopeful that 2018 will be the year that we move research out the lab and make it directly applicable to journalists debunking falsehoods.

If we want real-world solutions, we need real-world data, however. We need to learn how people actually consume fake news and fact checks when they encounter them in their everyday life, or when they look them up to prove (or disprove) an argument. The IFCN is coordinating fact-checkers to gather information about their audiences and we are eager to work with interested academics.

The biggest prize here, of course, would be data from Facebook. It’s been almost a year to the day since the social network launched a fake-news flagging mechanism, leaning on fact-checking organizations that are signatories of the IFCN code of principles. The platform has released extremely general figures about how this program has performed. I understand that the platform is under enormous pressure to make things work and that there are legions of Facebookers who genuinely care about fixing this problem. But more openness — and a willingness from academics and journalists to criticize constructively and not reflexively — has got to be the way forward.

Facebook has 2 billion monthly users worldwide. It drives around a third of all referrals to top publishers in the United States. And it is performing what is probably the largest real-life experiment on combating misinformation with fact-checking. We need to know how it’s going so that we can make better decisions about what fact-checkers should (and shouldn’t) do.