An effective description
of a numeric pattern should go beyond simply stating that two
variables are correlated or just reporting the values. It should
also interpret the size and shape of the pattern to show readers how
it relates to the question at hand. An effective way to do this is
to specify direction and magnitude of a pattern, for example stating
which category has the highest value and by how much it exceeds
other values of interest. Simple computations such as subtraction,
division, or percentage difference can be used to quantify the size
of the difference between values. See Miller (2004), Chapter 8 for
more on writing about quantitative comparisons.
