b 0 ¼ 146 : 2 ð 0 : 71 Þð 65 : 6 Þ
¼ 99 : 6For example, for a 50-year-old woman, it is predicted that her systolic blood
pressure would be about
yy^¼ 99 : 6 þð 0 : 71 Þð 50 Þ
¼135 mmHg8.1.5 Testing for Independence
In addition to being able topredictthe (mean) response at various levels of the
independent variable, regression data can also be used to test for the indepen-
dence between the two variables under investigation. Such a statistical test can
be viewed or approached in two ways: through the coe‰cient of correlation or
through the slope.
- The correlation coe‰cient rmeasures the strength of the relationship
between two variables. It is an estimate of an unknown population cor-
relation coe‰cientr(rho), the same way the sample meanxis used as an
estimate of an unknown population meanm. We are usually interested in
Figure 8.2 Graphical display for data in Example 8.2.SIMPLE REGRESSION ANALYSIS 289