Graphical data presentation.

Abstract

Figures and charts are the most influential vehicles for distributing scientific information, for affecting decisions as to the acceptance or rejection of a manuscript, and for attracting the attention of the scientific community to study results. Graphical excellence is mainly defined, first, by the highest possible data density (that is, the amount of information provided per graph area); second, by a low ink-to-data ratio (the avoidance of unnecessary shading, three-dimensionality, gridlines and what is often called 'chartjunk'); and third, by clear and unequivocal labelling of axes. The researcher's essential graphical toolbox should contain histograms, bar charts (always with measures of error), box-and-whiskers plots, scatter plots and forest plots.