Overview for Covariance

Use Covariance to determine the direction of the linear relationship between two continuous variables.

For example, an engineer wants to evaluate the direction of the linear relationship between hydrogen content, porosity, and strength of aluminum castings.

Covariance is similar to correlation but when the covariance is calculated, the data are not standardized. Therefore, the covariance is expressed in units that vary with the data and is not converted to a standardized scale of −1 to +1. Because the data are not standardized, you cannot use the covariance statistic to assess the strength of a linear relationship.

Where to find this analysis

To calculate the covariance, choose Stat > Basic Statistics > Covariance.

When to use an alternate analysis

To assess the strength of a relationship between two variables using a standardized scale of −1 to +1, use Correlation.
To check for nonlinear relationships graphically, create a Scatterplot or use Fitted Line Plot.