Correlation is a measure of the relationship
between two variables. The measure (identified by the variable
r) reflects both the strength of the relation on a
scale from 0 to 1 in absolute value and its direction - either positive or negative.
No relation is indicated when r is in the neighborhood of
zero. When | r | = 1, a perfect correspondence of values
exists.
A correlation can be displayed in several ways: a table of values,
a scatter plot, and, if the data has a geographic connection, in a
pair of data maps like those below. In each case you want to
look for patterns. In a table of values sort one of the
variables and look for a pattern in the other. In a scatter plot
look to see how closely the data conform to a line. And in a pair
of maps compare how the shading pattern in one map compares with
the other.
Notice in the maps below how the shading pattern in the X map is
practically the same has that in the Y map - indicating a strong,
positive correlation. Notice also how the values in the
table have been sorted on X and that the Y values are almost in
the same order. And finally, notice that the correlation
coefficient, r = 0.88 is a value fairly close to 1.
Experiment with the different types of relations indicated in the table. The values in the table will change automatically with each new selection. You will have to change the layers in the maps as indicated in the table.
Click to select the type of relationship. A Strong, Positive relation is selected initially.
1) Complete this sentence:
I know a strong positive correlation exists between two variables shown in a pair of maps when the shading patterns are ______________.
Write a similar sentence to describe each of the other possible types of relations:
- weak positive
- strong negative
- weak negative
- no relation
2) Which is the easiest type of correlation for you to see in a pair of maps?