How do people who have low trust in news sources decide which publications to trust? That’s the central question behind a newly published report from the Reuters Institute for the Study of Journalism’s Trust in News project.

The answer: People are quick to make judgments — snap judgments, as the people behind the study called them — when evaluating news outlets on popular digital outlets. These hasty decisions are based on a range of things that people look at, including the news brands themselves and who shared the stories.

To answer the question, the researchers polled 100 people in four different countries — Brazil, India, the United Kingdom and the United States — about their news habits.

Specifically, the authors chose participants labeled as “generally untrusting.” These volunteers were deemed as such because of their responses to the questions “How interested, if at all, would you say you are in politics?” and “Generally speaking, to what extent do you trust information from the following” list of 15 news organizations specific to their country. (In the U.S., this list included ABC, NBC News, Breitbart and others.)

Participants’ responses to these questions were measured on a five-point scale, and those whose scores suggested a below-average trust in news outlets as well as lower-than-average interest in politics were selected for the final sample. These people were also regular users of Facebook, WhatsApp, and Google.

With each of these participants, researchers conducted video interviews where the volunteers walked them through how they used each of the chosen internet platforms. This, the researchers write, “helped us observe in real time what they paid attention to in judging whether information was relevant and trustworthy to them. This technique allowed us to move beyond abstract responses about platform use to real-life experiences, where we could also probe participants further on specific and concrete examples.”

Here’s what they found:

These “generally untrusting” volunteers were unlikely to come across news on their regular platforms. When they did come across them, they were indifferent toward the news items. And the few times they did see them, the news tended to focus on softer topics such as entertainment.

When these participants did come across news articles on Facebook, Google or WhatsApp, they made quick judgments about the credibility of the information being reported. These judgments tended to be based on six main things, highlighted below:

When it came to headlines as a cue for making judgments, the researchers found that these people who weren’t attuned to news did get stopped by headlines — but the effect was perhaps the opposite of what outlets may have intended. One person is Brazil said, “The catchier the headline is, I’m more suspicious of it,” a sentiment that was echoed by another participant in the U.K., who said, “I think the more boring the heading is, maybe it’s more trust[worthy].”

The topic of the news item also played a role in how volunteers chose to trust publications. While these people tended to be skeptical of all news, they were especially skeptical of news about political topics. Here’s one U.K-based respondent’s take:

When you say “trust,” it depends. Trusting them for what? So, if I’m looking at a story about the floods down south, do I think they’re reporting that right? Probably. If I’m reading something about statistics that matter to politicians, do I believe it? No, because all the media are owned by the politicians.

What respondents paid attention to depended on the platform on which they saw news articles. On Facebook and WhatsApp, who shared the information helped inform how they viewed the news and the engagement the article was getting (likes, comments, etc.) Verification and labels on Facebook also helped. For example, one respondent in India said he trusted a news outlet “because this source has a blue tick, which means it’s verified through Facebook.”

At the same time, volunteers — much like the broader population — didn’t seem to know how platforms worked to show them news. They noticed that the source of the news wasn’t always apparent. People were also skeptical of the value to place on stories labeled as sponsored content. One volunteer in the U.S., for instance, said this of sponsored content on Google searches:

“Google is a private company. Google can be paid to be the first result you see. So, for certain subjects I would have to recall that it is very easy to pay to be in the first Google results.”

Participants also said they were concerned — rightly — that some of the social nature of these platforms (meaning friends and family were sharing news, which inherently made someone want to trust these people as sources) meant that it was easier to spread misinformation or mask dubious practices.

WhatsApp, for instance, offers more than just text messaging and news is often shared in audio format. But users expressed concern about this as well. One user from Brazil said this about her father’s use of audio on WhatsApp:

“[He] barely can read and write. He only uses audio messages, so news for him tends to be more trustworthy because he doesn’t know where it came from. So, it’s much more likely that he will believe in anything he receives from anyone.”

What does this mean for publications looking to win the trust of consumers?

“For news orgs, reaching this segment of the public may require more consistent and sustained branding efforts, in addition to tending more carefully to the precise ways in which stories are exhibited in digital spaces and how these may impact trust,” Amy Ross Arguedas, a postdoctoral research fellow at the Reuters Institute and the lead author of the paper said on Twitter.

And because these volunteers are coming across news on platforms that are not the news outlets’ own websites, the study “does put an onus on platforms to consider more carefully the role played by their design decisions and technologies in shaping users’ evaluations of news,” the authors write.

Read the full report here.