canonical correlation analysis
a method of data analysis that provides a measure of the strength of the relationship between a linear combination of predictor variables and a linear combination of outcome measures. Essentially, it extends a basic correlation analysis, which relates only one variable with another variable, and it extends multiple regression, which relates several independent variables with only one dependent variable. For example, a researcher might use a canonical correlation analysis to investigate whether several measures of attitude toward learning (e.g., confidence, anxiety, interest, enjoyment) are related to several measures of achievement (e.g., grade point average, the number of math classes taken, the number of memberships in honor societies).