PubMed2004
In some situations, tables, graphs, and figures can present certain types of information (including complicated relationships and sequences of events) more clearly and in less space than the same information would require in sentence form. However, do not use tables, graphs, and figures for small amounts of data that could be conveyed clearly and succinctly in a sentence. Also, do not reiterate in sentences the data that are shown in a table, graph, or figure: the point of creating a table or graph or figure is to eliminate that type of sentence from your manuscript. In building a data table you must balance the necessity that the table be complete with the equally important necessity that it not be too complex. Sometimes it is helpful to break a large table into several smaller ones to allow the reader to identify important information easily, but, conversely, it is a common mistake of novice authors to split up into several tables data that belong in one table. In almost all cases, only one table or graph or figure should be included in an abstract, and then only if it can convey essential information in less space and in a more easily interpretable way than the sentence form. For a poster, in almost all instances you should use only one typeface and one font in a table, graph, or figure. In general, do not use bold, italics, or color unless you are presenting a great deal of data and you need to highlight certain data values and you are certain that using bold, italics, or color will improve readability, which is rare. Do not include identical information in a table and a graph/figure. In reporting a clinical trial you will need to include a patient flow chart that identifies the number of patients initially screened for the study, the number of patients who were excluded (and why) after initial screening or in the final analysis, and how many patients entered, exited early, and completed each arm of the study. A treatment protocol should also be described with a flow chart. In preparing a graph the most common error is to include a line that suggests an unsubstantiated extrapolation between or beyond the data points. In selecting the graph's axes, avoid truncating, enlarging, or compressing the axes in ways that might make the graph confusing or misleading. To prepare clear, accurate, easily interpretable tables, graphs, and figures, rely on the rules described in authoritative guides such as the Council of Science Editors' Scientific Style and Format and the American Medical Association's Manual of Style.
Abstracting and Indexing↗