The correlation coefficient is a numerical measure of the strength of the relationship between two random variables. The value of the correlation coefficient varies from -1 to 1. A positive value means that the two variables under consideration have a positive linear relationship (i.e., an increase in one corresponds to an increase in the other) and are said to be positively correlated. A negative value indicates that the variables considered have a negative linear relationship (i.e., an increase in one corresponds to a decrease in the other) and are said to be negatively correlated. The closer the value is to +1 or -1, the stronger the degree of linear dependence.

Choosing a Correlation Test

Pearson's r Correlation

This widely-used coefficient measures the strength of a linear association between variables.

Spearman's Rank Order Correlation

The most common non-parametric measure, Spearman's is used when data are not normally distributed. Spearman's is a non-parametric equivalent of Pearson's correlation.

Kendall's tau Correlation

Another non-parametric method, used when analyzing data with one or more ordinal variables. Kendall’s is relatively "robust" to outliers.

Handling Missing Values

When there are missing values in your data, the Correlation Coefficient dialog provides the option to delete the cases pairwise or listwise.

Performing Correlation Coefficients

To open the Correlation Coefficients dialog box from the menu: