Many scholars are used to presenting their research data by means of mean standard deviation and histogram when writing papers and making reports, which has almost become everyone's habit, but this habit may have great problems. For example, sometimes we want to know the specific data behind these summary data. Sometimes we are talking about raw data, and you will feel completely ignorant, much like some movie actresses who use advanced makeup images, and the audience doesn't know what these stars are like. Or a photo after PS, you don't know what this person looks like. Finally, some scholars are committed to this problem, suggesting that magazines stop using this representation and provide their own tools to make scatter charts.

Academic circles have always been concerned about unreliable data analysis, and now an article in PLoS Biology questions the expression of statistical data. The article thinks that a series of data described by bar charts are often vague and misleading. It is time to stop this bad habit and let the author provide more comprehensive data details. This view was immediately widely supported on social media. Trevor Bedford, a virologist and computational biologist at Fred Hutchinson Cancer Research Center in Seattle, USA, shared the view that using bar charts to express data would blur the details of the data.

American Mayo Clinic scholars analyzed more than 600 academic papers published in the top journals of physiology in the first half of 20 14, and found that 85% of the papers used bar charts to describe continuous data. The author thinks that there is a big problem because the histogram does not contain the details of the data, but only provides the mean and dispersion (standard deviation and standard error) of the data. They use data to show that completely different data can be described by the same bar chart. They suggest that scatter plots should be used to represent data, especially for small sample studies. Only in this way can we truly reflect the details of the data and avoid misunderstanding of the data by readers.