Practical guide to choosing an appropriate data display.

The primary objective of graphing research data is to communicate key information visually in a rapid, accurate, and concise way. Graphs might be considered visual take-home lessons of the major point(s) of the manuscript. In choosing a graph, it is tempting to concentrate only on ways of illustrating summary statements characterizing the group(s). However, individual patients are unique, and their characteristics or outcomes may not be predicted by a group summary. Consequently, if possible, graphs should demonstrate individual responses as well as group summaries. "Graphical literacy," "graphical excellence," and "graphical acumen" are achievable with work and collaboration. To produce a well-designed graph, a combination of by-subject detail and overall results should be the goal within the same illustration. The practice gap addressed in this article is that little attention from authors, reviewers, editors, and publishers seems to be paid to graphical literacy. The purpose of this article is to present some practical guidelines for choosing or evaluating more appropriate data displays.