Write and position legends carefully

Legends (explanatory titles or captions) are a vital part of every figure, not a tacked-on afterthought. When readers turn a page, they look first at illustrations, then at the legends. Each title should orient the readers toward the figure's meaning and enable them to identify its components without referring to the text. It should also differentiate that particular figure from all other illustrations in the paper.

Like a table title, a figure legend should be brief - 8 to 12 words, as a rule of thumb - and need not be a complete sentence. A good legend should contain only enough detail for readers to understand the illustration; consign the necessary minutiae to the text.

Too vague: Fig. 1. Graph of Relevant Data.

Over-specified: Fig. 1. Outcome of Multifactorial Analysis of Relationship Between Symptoms, Chronology of Appearance, Diagnostic Signs, Blood Work Constants From The Literature, Health Outcome, and Other Parameters For Selected Group of Fifteen Adult Ostriches.

Better: Fig. 1. Multifactorial Analysis of Health Records of Fifteen Adult Ostriches.

For a classic research paper, the legend for a figure should never be included within the figure itself. In the final version of the typescript, list the figure legends (double-spaced) on a separate page at the end of the text. For a verbal presentation, place the legend directly on each figure.


Tables present results; graphs promote understanding of results and suggest interpretations of their meaning and relationships. Most graphs are based on a set of numbers, just as most tables are. But, because graphs are fundamentally pictures rather than a set of numbers, information generally is easier for the reader to grasp than if it were printed in columns.

Consider graphing data when you feel that the relationships are more vital to your message than the actual numerical values themselves. Thus, you might use a graph to present trends dealing with two related variables, one or more variables changing through time, or data interesting for the magnitudes of differences which might be related to unknown factors or experimental manipulations.

A graph always shows how one parameter varies relative to changes in another. One factor may be controlled and varied (temperature, for example) while some effect of this change is measured. Alternatively, the effect of changes in some uncontrollable factor such as time may be measured. Temperature and time in these examples are independent variables, which are usually plotted in relation to the horizontal (x)-axis. The effect these changes have on something else (the dependent variable) is normally plotted in relation to the vertical

10 000


f 6000

' 1U.S.A.

c o o


o 4000



^ 2000




i 1


1 1


1989 1990 Year



Fig. 3.2. This effective line graph shows changes in rabies disease incidence over time.

Fig. 3.2. This effective line graph shows changes in rabies disease incidence over time.

A range of software programs can produce sophisticated graphs with the touch of a finger. Use them initially to construct various graphic presentations of the data and consider the alternatives. Then modify such aspects as fonts, type size, shadings, and symbols to produce visually dynamic illustrations. Remember, however, that legibility and comprehensibility should remain the most important criteria. For further help, see Robbins (2005).

Was this article helpful?

0 0

Post a comment