Add connecting lines between barsegments to facilitate comparisons.Line Graphs Line graphs are used to portray continuous movement or change in a measured
variable, typically against a time
scale. They are best for showing the
trend in a set of data over time, but are
less effective in comparing amounts of
two variables.
· Five or six lines on one graph are
usually plenty. The number that can be
presented clearly on one graph depends
on the position of the curves and the
function of the graph.
· Use thick, solid, colored lines for
projection and patterned lines for blackand-
white printouts
· Use a different color for each line,
making the most important line the
lightest and brightest color.
· Data lines should be the heaviest
lines on the graph.
· Data points should always stand out
from and dominate the connecting linework,
since the data points are the actual
data and the connecting lines are usually
just a visual device to show trends
and continuity.
· Symbols and connecting lines
should dominate all other elements of a
chart.
· Symbol diameter should be about
two to three times the width of any connecting
lines.
· Polygon symbols should be used as
markers or symbols. Recommended
symbols are filled circle, triangle, and
square, and open circle, triangle, and
square. Other symbols are not as distinctive
and, if used, should be carefully
oriented, so that they can be distinguished
from the recommended ones
and from the connecting line. Avoid using
X’s, crosses, or symbols with dots.
· For data points that overlap, the
symbols should be drawn overlapped.
· All symbols should look the same
size.
· Curves should not extend before the
first data point or after the last data
point.
· Identify curves with a brief label
next to them. Place the label close to the
curve. It should be contained within the
rectangle implied by the axes and should
be oriented horizontally.
· Don’t use arrow or leader lines, as
they clutter the graph.
Figs. 28 and 29 illustrate the above
points, with Fig. 29 being better.
Pie Charts or Stacked Bar Charts
(Component Bar Charts)
Pie charts illustrate the component
parts or percentages of a whole, where
the values must add up to 100%. Pie
charts are the least efficient means of
charting small data sets and are not often
used in the sciences, but are common
in business and general information
graphics.
Stacked bars are often used as spaceefficient
replacements for multiple pie
charts. Neither pie charts nor stacked
bar charts are impressive if too many elements
are displayed. About six segments
would be the limit before visual
confusion sets in.
· Use a maximum of six slices in a pie
chart. If you need more, group smaller
ones into one group titled “miscellaneous”
or “other,” and pull this group
out into another chart.
· Order slices beginning with the
largest slice or the most important data
starting at the 12 o’clock position and
arrange subsequent slices clockwise. The
most important slice is in the upperright
quadrant. Arrange slices in order
from largest to smallest, clockwise.
· Emphasize one slice by “exploding
it” or moving it out of the circle. You
can also choose a color or fill pattern
different from the rest of the slices to
emphasize a particular slice.
· For stacked bar charts, labels are
placed to the right of the vertical graph
or above a horizontal graph.
· For pie charts, labels are placed inside
the segments if the labels are brief
and segments are large enough. Otherwise,
labels are listed next to the segment.
Labels should either be in or out
of the chart; don’t mix label styles (Figs
30 and 31).
· Percentages should be included in
labels. Use whole numbers.
· Percentages are usually separated
from the labels (on a separate line) and
are smaller in size than the labels.
· Labels and percentages should be
placed horizontally, not at an angle
· Each slice should be easily distinguishable
from the rest and clearly labeled.
Area Charts
Area charts emphasize the volume
of data, from the baseline of a chart to
the top of a trend line. There are two
types of area charts, cumulative and
stacked:
· Cumulative charts consist of lines
on a multiline chart with the areas beneath
each line shaded. Each line is
measured from the baseline.
· Stacked charts also show several
data sets, but each area is measured
from the previous area. Areas do not
share a common baseline. Because they
don’t share a baseline, order of the