How to reliably report imprecise data. A meta-analysis by the Max Planck Institute

illustration: DALL-E

Journalists describe phenomena every day that science has not fully explained. Modern media reports on health, climate, or technology very often contain elements of risk and uncertainty. Charlotte Dries and her team from the Max Planck Institute for Human Development attempted to systematize knowledge about this phenomenon.

The researchers published a report in the journal Royal Society Open Science titled The impact of uncertainty communication on trust in its sources: a systematic review and meta-analysis. The authors analyzed the results of experiments testing whether open communication about a lack of knowledge destroys public trust in the source. Many publishers fear that admitting ignorance will undermine editorial credibility. The authors of the report checked whether these anxieties are justified in light of empirical evidence.

• Scientists conducted 27 out of 28 experiments using fictitious news texts.
• The authors of fifteen selected studies did not specify a concrete source for the message, which made recipients evaluate the press content itself.
• Exactly 12 research projects used the form of a classic article from a newspaper or news portal.

The absence of a specified author makes readers focus their attention directly on the linguistic structure and the facts that the authors included in the text. This anonymity reflects the situation where media consumers read random reports on the internet.

We do not trust text without an author

For example, when a person browses social media and encounters an article about a new biological threat, they often do not know the reputation of the specific author. At that point, the editorial phrasing determines how deeply the recipient will believe the presented arguments. The team led by Charlotte Dries proved that the public processes such information very cautiously, without showing extreme defensive reactions to news about gaps in expert knowledge.

Message sender	Number of experiments
Scientific community	7
Government institutions	6
Media newsrooms	4

These numbers show that media researchers most frequently test the authority of science, while they less often look at the intermediaries of information. Journalists rarely generate primary medical knowledge; they usually quote statements from experts or summarize public reports. Charlotte Dries and her colleagues note that this structure of news naturally dilutes responsibility for lack of precision. The newspaper reader can transfer potential grievances to research laboratories, sparing the newsroom itself, which reliably reports the actual facts. This psychological mechanism protects the position of the news media as an objective mirror of reality.

The methodology of the collected studies revealed problems with maintaining experimental purity in communication analyses.

• Scientists reveal that 18 out of 28 experiments contained hidden uncertainty in control conditions. This means that texts that were supposed to pass for certain used words such as about or maybe.
• Additionally, in 9 studies, scientists detected confounding factors in the interventions themselves, where authors modified the complexity of the text or added information unrelated to the topic.

Despite these imperfections, the overall market picture remains stable and does not show drastic fluctuations in media credibility ratings.

The presentation of data and reader reactions

Choosing the appropriate data presentation format represents a daily challenge for every editorial team. The authors of the report describe how numerical and verbal forms affect the psyche of the reader. Journalists can present a lack of certainty using numerical ranges, for example, by writing about an increase in unemployment of 5 to 7 percent. An alternative path is using descriptive phrases, such as a probable increase or slight fluctuations. Experiments show that dry numbers enjoy greater tolerance from the public than ambiguous verbal descriptions, which can arouse anxiety.

• A precise mathematical formulation builds an image of professionalism in the eyes of recipients, even if these numbers communicate a lack of final resolutions.
• When a reader sees a specified confidence interval or margin of error, they subconsciously attribute high reliability and methodological accuracy to the newsroom.
• Words tend to be more ambiguous and can suggest that a journalist deliberately avoids specifics or does not possess sufficient competence to describe the topic.
• For example, a press release saying that a new drug shows some efficacy sounds suspicious, while information about an efficacy of 80 percent with a margin of error gives a sense of objectivity.

Scientists from the University of Cambridge and the Max Planck Institute confirm that numbers have a calming effect on news consumers. The researchers also emphasize that only a few studies included modern visual forms, such as infographics or charts, which dominate contemporary news portals. As many as 16 papers relied exclusively on verbal messages, and only 9 combined words with numbers. This shows that theoretical knowledge about media communication still does not keep pace with the real practice of graphic design in the modern press.


Journalists intuitively use progress bars or weather maps with marked risks, while science still mainly investigates traditional, linear text messages. This discrepancy places an urgent need before future researchers to verify how modern visual society reacts to graphic charts of uncertainty.

Challenges in building relationships with the message recipient

The most optimistic conclusion from the collected data concerns the strategic safeguarding of public trust in crisis situations. Experiments demonstrated that open communication about the variability of knowledge acts as an effective buffer against later accusations of lying. When media loyally inform from the beginning that data are preliminary, a subsequent change of facts does not provoke reader anger. For example, during a virus epidemic, dynamic changes in medical guidelines can destroy the reputation of authorities if they previously pretended to be absolutely infallible. If, however, press articles mentioned the dynamic nature of discoveries from the start, the public shows great understanding for corrections.

The psychological analysis of readers also brought surprising conclusions regarding the demographic and ideological characteristics of the public. Factors such as partisan sympathies, general education, intelligence level, or individual tolerance for ambiguity do not change the way these messages resonate. People with radical political views process information about a lack of certainty in exactly the same way as individuals with centrist views. Newspaper newsrooms do not need to adapt the degree of openness of their texts to the ideological profile of their subscribers. The only significant factor modifying reactions turned out to be the alignment of the presented evidence with the prior personal beliefs of the given reader.

The main barrier to fully translating these results into daily editorial practice remains the ephemeral nature of the conducted tests. Almost all experiments measured trust immediately after reading the prepared text, without offering any long-term studies. Readers build a deep relationship with a newspaper over years, when the newsroom delivers reliable analyses and commentaries to them daily. A single contact with a text containing scientific caveats might not cause an immediate change in attitudes, but regular exposure to uncertainty could tire the recipient. Modern newsrooms must balance independently between scientific honesty and the natural need to provide recipients with clear, useful guideposts in a complex world.

The full article "The impact of uncertainty communication on trust in its sources: A systematic review and meta-analysis" is available at https://pure.mpg.de/view/item_3698296