Layout Design 43
that people can understand on a poster‑sized presentation. For example, I have seen
the advice, “Don’t insert long tables of numerical data or text” on conference poster
guidelines. I think this advice comes from underestimating the human brain’s ability
to discern patterns in large amounts of data. If we format the table properly we can,
indeed, present long tables, and we should if they further the message that we are
trying to get across. This also increases your map’s credibility.
Best Practices
If you have a table showing many attributes for the same features, you could alternatively
choose to create a series of maps with each one showing the features and one
particular attribute. This is a judgment call based on the space available, the size of
the features, and the complexity of the attributes.
Style
There are many table styles. The important guidelines are to ensure that the column
and row headings are legible, that the data line up, and that the column and row lines
do not detract from the data. Because the map is still likely to be of more importance,
the table ought to include as few ornamental lines and borders as possible to ensure
that it has less visual weight than the map. In fact, delineating every cell in a table with
a border usually adds unnecessary clutter, and you may also consider doing away with
any lines at all and instead using white space as the main separator, with perhaps one
or two simple neat lines separating the table from the elements above and below it.
To further enhance the table, you can add visual weight to certain elements.
Anomalies in the data that may be too hard to detect on the map can be highlighted
in the table with a circle or bold text. These enhancements need to highlight only
the pertinent data cell, not the whole table row. The conservative use of color — to
differentiate assets (black) from debt (red), for example — can also add to a viewer’s
understanding of the data. Another trick is to change the locations of numbers based
on some characteristic of the numbers. For example, a table of acres of impervious
surface per watershed could show the impervious acreages at a distance from the
watershed name in proportion to the watershed’s size. This would be a great accompaniment
to a map of watershed‑level impervious surface normalized by watershed
area because it would be presenting the viewer with the unnormalized numbers in
the table but generalized percentages in the map (see Figure 3.19).
Fish Creek 25|150
Barberville 150|180
Granton 50|200
Lakeland 62|250
Upwater 124|320
Figure 3.19 Tables can be formatted to clearly portray patterns in the data. In this
example, the distance between the ratios and the place names increases proportionally to
the denominators. When the data are presented like this it is easier to see that Barberville’s
numerator does not follow the pattern. The circle highlights this anomaly.