most important principles are:
◆ Always have completely informative headings and labelling,
including full details of units or measurement and what
the cell contents show.
◆ Use the ‘need to know’ criterion to pick an appropriate level
of detail for numbers. Choose the minimum number of
decimal points needed. If you make use of index numbers or
ratios, choose levels which give the most easily
understandable numbers for readers. Consider how many
‘effective digits’ are needed, and use rounding or number
simplification appropriately.
◆ Design all tables to show a numerical progression (exceptfor
tables showing over-time trends or categorical variables with
a fixed order).
The final issue to consider about tables is whether you really
need them at all. Would it be better to use a chart or graph
instead of a table? In most cases charts will be better because
they are clearer and more visual. Tables should principally be
retained for the following circumstances:
- There are only a small amount of data to present, so that a
simplifying chart is unnecessary. - Readers need to know numerical values more precisely than
would be shown in a chart, for instance if there are fairly
small variations in results. - The data to be displayed have very strong variation between
the lowest and highest numbers so that it would be difficult
to display the range of the data effectively in a chart.
For instance, isolated high numbers for one or a few years in
an over-time chart might necessitate a scale which would
mean that readers could not detect any visible differences in
other years’ figures, whereas in a table they could still be seen. - You want to compare data scaled in very different kinds of
units or indices, and they could not easily be accommodated
on one chart. Alternatively you might have the numbers in
different columns which are of the same kind, but of such
different sizes that they would be hard to scale together on a
graph. Here tables can save space, since otherwise you
HANDLING ATTENTION POINTS◆ 171