148 Fundamentals of Statistics
This stem and leaf plot is at a disadvantage compared to the histogram
in that the size of each bin is directly determined by the data values them-
selves. Stem and leaf plots also don’t work well for large data sets where
each stem will need to display a large number of leaves. One way of modi-
fying a stem and leaf plot is to split the stems into subgroups. For example
you can split a stem into two groups: those with leaves having values from
0 to 4 and those with leaves from 5 to 9. Doing this for the above chart yields
the following stem and leaf plot:
1003 |
1 | 2
1 | 8
2 | 344
2 | 79
3 | 12
3 | 5
4 | 11
4 |
5 |
5 | 5
6 |
6 |
7 | 1
7 |Another modifi cation to the stem and leaf plot is to truncate lower and
upper values in order to reduce the range of stems in the plot. This is use-
ful in situations where you have an extreme value whose presence would
greatly elongate the plot’s appearance. For example, if the value 2,420 is
added to the above data set, then the resulting stem and leaf plot will have
a long stem with a long list of empty leafs. In this case, removing this value
from the stem and leaf plot, but noting its value elsewhere, might be the best
course of action. The plot might look as follows:
1003 |
1 | 28
2 | 34479
3 | 125
4 | 11
5 | 5
6 |
7 | 1
2400Excel does not have a command to create stem and leaf plots, but you can
create one using StatPlus. Let’s create a stem and leaf plot for the home price
data and compare it to the histogram we created earlier. As before, we’ll break
the stem and leaf plot down using the values of the NE Sector variable.