During an impassioned appeal on "Reliable Sources" today, CNN senior media correspondent Brian Stelter called on journalists to provide "forceful rebuttals" to conspiracy theories espoused by Republican presidential frontrunner Donald Trump.

The country's top fact-checkers have been doing so for months. Factcheck.org crowned Trump last year 'King of Whoppers,' an uncharacteristically strong choice of words for a website that has long avoided using irreverent ratings in its fact checks.

Stelter was, however, primarily addressing TV journalists who interview Trump on air. His call comes a day after Washington Post fact-checker Glenn Kessler noted that "astonishingly, television hosts rarely challenge Trump when he makes a claim that already has been found to be false."

Yet even when they try, it doesn't seem to change Trump's behavior. Stelter replayed clips of Sunday show hosts Chuck Todd and George Stephanopoulos pushing back on Trump's falsehoods earlier in this campaign. His reaction to being told he was wrong? "It was well-covered at the time, George," and, "All I know is what's on the internet."

The reality is that how TV hosts check facts on air is as important to changing a politician's behavior and a viewer's mind as whether they fact-check at all.

So if, as I hope, more interviewers decide to pick up Stelter's challenge, they should keep in mind three things about fact-checking live on air (beyond being very, very prepared).

First of all, the fact check should frame a claim as being false clearly and succinctly. Replaying it with excessive prominence and no accompanying indication of its falsehood could lead people to confuse the claim for news, especially if they're watching in a noisy place or turning on the television after the claim was introduced.

This was not necessarily what happened in Jake Tapper's otherwise powerful takedown of Trump's allegation that Cruz' father was in a photo with John F. Kennedy assassin Lee Harvey Oswald. During the segment, he very prominently displayed Trump's quote for 20 seconds, which could be confusing to viewers. As noted in a 2012 review of fact-checking formats from the New America Foundation, "people may use the familiarity of a false claim as a heuristic for its accuracy."

Secondly, fact-checking on TV should use images and graphs. If it boils down to the host's word against the candidate's, viewers may choose to believe whomever they trusted more to start with. Instead, a TV fact-checker should be presenting sources prominently and visually. Fox News illustrated this in March, when it countered Trump's flawed budget proposal with graphics that broke down federal spending figures.

To stay with Trump's falsehoods, he has asserted that unemployment is actually over 40 percent rather than approximately 5 percent. An effective way to reinforce the scale of this falsehood is to project the unemployment figures and indicate clearly the indicator is a reliable source. When I was a fact-checker on Italian TV, I worked closely with producer Alberto Puoti, who constantly stressed that the strength of graphics and the transparency of sourcing is what separates fact-checkers from talking heads, fact checks from opinions.

Finally, fact checks should provide alternative explanations. If that man with Lee Harvey Oswald wasn't Rafael Cruz, who was it? Why could Rafael Cruz not have possibly been in that photo? Even if someone has been convinced that a claim is false, the most effective way to keep them from remembering that explanation is if they have a different one than the misinformation being peddled. The Debunking Handbook provides some tips for doing so.

Fact-checking on TV in the United States has yet to be expressed in formats like those attempted in, say, Australia or Spain. Until that happens, TV fact checks will not be as effective as truth sleuths would like them to be.