ABSTRACT
The simple t-test and the confidence interval for a difference in means (described in Chapter 2) compare the measurements for an outcome of interest in two different groups. Often, however, researchers are interested in comparing results between sev eral groups, or determining the associations between an outcome and other con tinuous variables. The correlation coefficient (see Chapter 2) measures the linear relationship between two variables, ranging from 1 (a perfect positive linear rela tionship) through zero (no linear relationship) to -1 (a perfect negative linear rela tionship). However, while it provides the strength o f the association, it cannot be used for prediction and is restricted to linear relationships between two variables.