Figures, percentages and calculations, the media loves to use such data that contain (simple) messages that can be used to create newsworthy stories. News organisations and their audiences are very interested in statistics, especially when they seem to be related to our daily lives (Wormer, 2007). But with the use of statistical data for news comes a risk as well, and that is telling the wrong story. Of course you can discuss the severity of wrong interpretations in statistical data, but the fact remains that one may be misinformed about serious topics. This phenomenon has recently appeared in various news media.
Why bacon seemed to be dangerous
In october this year, the World Health Organisation (2015) announced that processed meat, such as sausages or bacon, can cause cancer. The International Agency for Research on Cancer (IARC) has surveyed over 800 studies on the issue and listed processed meat in the same category as smoking, alcohol, uranium, and exposure to solar radiation (Stats, 2015). The IARC also found that red meat was probable carcinogenic to humans. After these announcements from the WHO, many news organisations started to report about the findings of the study from the IARC: The NOS, the NBC, the AD, the FD, the NRC, the Telegraaf and many more reported about the carcinogenic risks of processed meat. Although this story went “viral” in the news, it appears that not many media sources understood the real implications of the study.
The increase of absolute risk or the risk we already had… Who cares!?
In some of the articles that were published in the news, the media reported about the increased risks of getting cancer by eating processed or red meat. For instance the NOS, Nu.nl, and the Telegraaf reported that people who eat 50 grams or more on a daily basis, would have an increased risk of 18% to get colorectal cancer compared to people who eat less meat.
Now this reasoning was not exactly correct, because this was not what the IARC meant with their publications. When the IARC argued that ‘each 50 gram portion of processed meat eaten daily increases the risk of colorectal cancer by 18%’, they actually meant an increase in risk that we already had (which is less than 18%). To illustrate this, if for instance the risk to get cancer is 10%, an increase of 18% in a 10% risk results in an increase of 1.8%. This means our absolute risk would increase from 10% to 11.8% (Stats, 2015), which is less than the some news organisations reported. The problem is that these findings are less dramatic than the 18% which was eventually mentioned. But there were new organizations that went even further.
Some media, as for instance the Metro and the Daily Mail, reported that the consumption of processed meat was just as carcinogenic as smoking cigarettes. But an infographic from Cancer Research UK (2015) shows that this is absolutely not the case. Although significant results were found that both processed meat and smoking cigarettes causes cancer, does not mean the risk and damage are the same. There is a far greater risk of getting cancer through smoking cigarettes, and the number of cases could be much more reduced by quit smoking than by stop eating meat.
So what does this mean for (data) journalism?
Now this is may seem just one example of a recent misinterpreted breaking news story, but this happens more often than we think. Not only in the news, but also in science there is an attitude of the most dramatic findings are also the most likely to get published (Lehrer, 2010). In science, opportunistic biases or invalid findings has negatively affected the way that the public regards research and scientists. People tend to doubt about the validity of research findings because researchers are suspected to be motivated by political, economic or social agendas (DeCoster, Sparks, Sparks, Sparks & Sparks, 2015). I believe the same thing can happen for journalists and news organisations. However, with the emerge of big data and data journalism, more and more people have accessibility to data sets, and therefore more and more people can act as a journalist. This leads to greater importance of statistics and statistical reasoning in journalism, because the availability of data will give people the opportunity to check everything they read in the news (Nguyen & Lugo-Ocando, 2015). There is a risk that people will increasingly mistrust journalism and the media, if it turns out that the stories they publish are not always based on truth. And with the continued growth of data and analytical tools, we may be close at a point in the future we can reveal how often news organisations tell us the truth or lie about breaking stories.
- Cancer Research UK. (2015). Processed meat and cancer – what you need to know.
- DeCoster, J., Sparks, E. A., Sparks, J. C., Sparks, G. G., & Sparks, C. W. (2015). Opportunistic biases: Their origins, effects, and an integrated solution.American Psychologist, 70(6), 499.
- Lehrer, J. (2010). The truth wears off: Is there something wrong with the scientific method? The New Yorker, 13, 52 – 57.
- Nguyen, A., & Lugo-Ocando, J. (2015). The state of statistics in journalism and journalism education: issues and debates. Journalism: Theory, Practice & Criticism, 15(7).
- Stats. (2015). Death by bacon: did the news get to the meat of the matter.
- Wormer, H. (2007). Figures, statistics and the journalist: an affair between love and fear Some perspectives of statistical consulting in journalism.