If interpreting statistics or results of research is often complicated for health professionals, the task can be much more arduous for journalists involved in scientific information. To shed a little light on some basic concepts and to discuss the problems which biomedical reporters usually face when understanding statistical data, the Session on biostatistics for journalists and communicators took place, the second one to be held of this nature, on 10 June 2015, at the Faculty of Philology, Translation and Communication of Universitat de València, in collaboration with the Spanish Association of Scientific Communication.
The first paper was presented by Gonzalo Casino, author of Escepticemia.com and coordinator of the session. As a scientific journalist, he went through some of the most common errors within the profession, such as ignoring the absolute risk, mythicizing prevention by default, or paying too much attention to animal studies. According to Gonzalo Casino, it is essential for journalists to stop using the pet phrase “according to the study”, without specifying the type of work and the confidence it merits.
The person responsible for clarifying statistical concepts was Erik Cobo, of the department of Statistics and Operational Research of Universidad Politécnica de Catalunya. Research or development? Intervention or prediction? Experiment or observation? Cause or effect? These were some of the terms that the author of Biostatistics for non-statisticians wanted to differentiate. He also underlined the importance of the interval of confidence above the P-value. He concluded that more than size, the quality of the sample should be corrected to assess the relevance of a study.
In the pyramid of scientific evidence, systematic revisions occupy the apex. In principle, they are the works that generate more trust and reliability. However, according to Pablo Alonso, researcher of the Cochrane Collaboration, the international organization that promotes this type of studies, two elements should be considered to assess the quality of revisions: that there is an explicit search in various data bases and that the quality of analysed studies is assessed.
The three papers on biostatistics were complemented by two practical workshops, in which participants, the majority scientific journalists and communicators, worked on articles, press releases and real news articles. The session ended with a debate with Carolina Moreno, professor of scientific journalism of Universitat de València; Ignacio Fernández Bayo, vice-president of the Spanish Association of Scientific Communication, and Jaime Prats, journalist of the newspaper El País and specialized in health and biomedical information.
All contents of the Session on biostatistics of journalists and communicators are available in the Esteve Foundation Notebook.