46 FREQUENCY TABLES AND THEIR GRAPHS
Figure 4.5 Distribution curves.to the right, and is common in biomedical and other data; (c) is called bimodal since
it has two distinct high values with a dip between them; and (d) has a long tail to the
left or is called skewed to the left. Forms (c) and (d) are less common in practice.
The frequency distribution of the population is an important concept. Actually, we
rarely know the exact form of the distribution of the population we wish to study. We
have a sample from the population, and we plot a histogram (or frequency polygon)
from the sample; this can be regarded as giving some idea of the form of the frequency
distribution of the population. One important use of the histogram or stem and leaf
graph is to check visually if our sample has a form that is similar to what we expected
in the population. These plots are easy to obtain from the statistical software packages.
There are numerous other methods of depicting data graphically. Texts on this
subject include Cleveland [1985], Cleveland [1993], and Chambers et al. [1983]. For
a discussion more oriented to the visual impact of the graphics, see Tufte [1990].
Tables are also an obvious way of presenting data, and data are often presented
in that form. Koschat [2005] gives some suggestions for making tables. One is that
the first decision is what variables to make rows and what variables, columns. The
first rule is that, in general, numerical comparisons are easier to make within columns
than within rows, so if the reader wants to compare ages of patients, the ages should
be kept in a column. Put the rows whose entries you want to compare close together;
the same holds true for columns. Since we tend to read from left to right and top to
bottom, the reader will probably look first at the upper left corner of the table.