In a crisis, people believe what they see more than what they are told

Ton (a fictional character) heads communications at a public agency. During an emergency his team issued a statement every day, accurate in every line. Yet the more statements they put out, the more people on social media questioned them. He was not sure whether the problem was the wording or something else.

The three studies brought together here look at communication in very different situations, but they reach the same lesson: in a crisis, people judge by what they see happening, more than by what they are asked to believe.

Belief rose and fell with visible results

A study by Assoc. Prof. Dr Pavel Slutskiy and Assoc. Prof. Dr Smith Boonchutima (published in 2022) analysed Facebook and Twitter posts and comments from early 2020 to early 2022, alongside interviews with 50 experts. While the outbreak was under control, people barely questioned official information, and mask-wearing reached almost 95% before the World Health Organization was even recommending it. But in the third wave of 2021, when April alone brought more new cases than the country's entire previous total, belief went the way of performance.

The researchers conclude that credibility in that period depended on whether the government seemed to be winning, more than on how true or transparent its messages were. When people stopped believing, they did not stop looking for information; they turned to international bodies instead. And a rule restricting information that caused public fear, even when it was true, read as silencing rather than solving.

Trust can shift faster than we expect, and not only trust in government. An online survey by the Reuters Institute, published via Statista, shows overall trust in news among Thais hovering around half for years, then falling from 55% in 2025 to 47% in 2026, the lowest in the six years of data.

If the organisation does not call it a crisis, no one owns it

A study by Theekakul and Asst. Prof. Dr Teerada Chongkolrattanaporn (published in 2021) looked at three digital TV stations penalised by the regulator over their coverage of the 2020 hostage-taking in Nakhon Ratchasima. Only one communicated online, with an apology on Facebook about two days after the incident ended. The post drew more than 3,200 comments, mostly critical, and the station replied to none of them.

The finding that applies to any organisation is that none of the three stations saw the penalty as a crisis for itself, so there was no plan, no manual and no one assigned in advance. The newsroom was left to handle it, even though the newsroom was a direct party to the dispute. The specialists interviewed recommended putting corporate communications in charge, setting up a war room with thresholds agreed beforehand, and apologising in full.

People pass on facts more than feelings

A study by Asst. Prof. Dr Teerada Chongkolrattanaporn (published in 2018) analysed the 2013 Twitter campaign against the Mae Wong dam. The campaigners told the story of a 388-kilometre walk in emotional language, but the messages people actually passed on were the ones with facts and evidence for cancelling the dam. When a documentary was pulled from broadcast, the team put it on YouTube instead, and the petition on Change.org became the largest in Thailand at the time.

Campaigners, the news media and the public each framed the issue differently, yet the campaign worked. The responsible minister sent a representative to receive the petition and said the impact report would be examined carefully and construction would not be rushed.

Four points to prepare for the next crisis

1. 1

   Agree in advance what counts as a crisis

   Write the thresholds down beforehand, including issues a regulator treats as a problem. If the organisation does not count it as a crisis, no one will be assigned to it.

2. 2

   Put someone who is not a party in charge

   Give the work to corporate communications or a central team, not the unit at the centre of the story, which tends to explain its own procedures instead of apologising.

3. 3

   Say plainly when results are not good yet

   When what people see contradicts what you say, more explanation does not help. Say what is being fixed, and cite sources people trust, such as international bodies, rather than competing with them.

4. 4

   Prepare both facts and story

   Stories draw attention, but what people pass on tends to be checkable fact. Prepare short messages with figures and evidence that are easy to share.

Reading this evidence with care

The COVID study is a qualitative analysis. It reports no post counts or proportions of opinion, so the link between approval and credibility is read from the sequence of events rather than measured, and the approval figures come from polls and news reports rather than the researchers' own data. The TV study covers one event, three stations and seven interviewees. The Mae Wong study is a single case from 2013 data, mainly from Twitter, and judges success by the policy outcome rather than by measuring attitude change directly. The Reuters Institute news-trust figures come from an online survey, so they mainly reflect internet users, and they measure trust in news overall rather than trust in government. This article deliberately does not name the individuals, TV stations or companies criticised in the original studies, because the lesson lies in the mechanism, not in who was involved.

Credibility is built before a crisis and proved during one

In a crisis, good wording does little when what people see points the other way. The organisations that come through are the ones that prepared people and thresholds in advance, admit it when they get things wrong, and give facts people can check and pass on.

If your organisation is planning how to communicate in a crisis, the Center of Excellence in Communication Innovation at Chulalongkorn University's Faculty of Communication Arts would be glad to talk and share what we have learned.