Bar graphs (Fig. 3.3), also sometimes called column graphs, are used to present discrete variables in a visually forceful way. They are a single axis graph used
to compare size and magnitude of discontinuous data. They are superior to circle and line graphs for showing relationships, magnitudes, and distributions (Macdonald-Ross, 1977b). On the negative side, bar graphs generally provide a relatively small amount of information while taking up a fairly large amount of space.
The bars may run either vertically or horizontally, but are most effective when they run in the direction in which people expect to see them. Thus, vertical bars are usually used for such data as temperature and weight, horizontal bars for distance, time, and speed. Whatever type of bar graph you choose, make the bars the same width, and the space between bars or bar groups one-half of a bar width.
Subdividing bars by shading or cross-hatching adds another dimension of information. Divided bar graphs, for example, can compare percentages of a whole rather than relative size. In this, they function much like pie charts but are less effective. Superficially similar to bar graphs, histograms are two-axis graphs that typically show frequency distributions by use of a series of contiguous rectangular bars.
Was this article helpful?