An attribute with a value of z1z2 is integrated into the standard linear regression model to account for two predictive variables, z1 and z2, and their impact on the outcome variable.
What is this kind of effect referred to as?