The CORR Procedure

Overview

The CORR procedure is a statistical procedure for numeric random variables that computes Pearson correlation coefficients, three nonparametric measures of association, and the probabilities associated with these statistics. The correlation statistics include

Pearson product-moment and weighted product-moment correlation
Spearman rank-order correlation
Kendall's tau-b
Hoeffding's measure of dependence, D
Pearson, Spearman, and Kendall partial correlation.

PROC CORR also computes Cronbach's coefficient alpha for estimating reliability.

The default correlation analysis includes descriptive statistics, Pearson correlation statistics, and probabilities for each analysis variable. You can save the correlation statistics in a SAS data set for use with other statistical and reporting procedures.

Simple Correlation Analysis for a Fitness Study Using PROC CORR is the simplest form of PROC CORR output. Pearson correlation statistics are computed for all numeric variables from a study investigating the effect of exercise on physical fitness. The statements that produce the output follow:

options pagesize=60;
proc corr data=fitness;
run;

Simple Correlation Analysis for a Fitness Study Using PROC CORR

Customized Correlation Analysis with Partial Covariances and Correlation Statistics and Output Data Set with Pearson Partial Correlation Statistics illustrate the use of PROC CORR to calculate partial correlation statistics for the fitness study and to store the results in an output data set. The statements that produce the analysis also

suppress the descriptive statistics
select and label analysis variables
exclude all observations with missing values
calculate the partial covariance matrix
calculate three types of partial correlation coefficients
generate an output data set that contains Pearson correlation statistics and print the output data set.

For an explanation of the program that produces the following output, see Storing Partial Correlations in an Output Data Set .

Customized Correlation Analysis with Partial Covariances and Correlation Statistics

Output Data Set with Pearson Partial Correlation Statistics

Chapter Contents
Previous
Next
Top of Page