If I set up an interaction term between two dichotomous variables, let's say experiment (1= control, 2 = experiment group) and gender (1 = male , 2 = female). If the effect of the interaction term (experiment*gender) on happiness is .46 using regression, does that mean the experiment increased happiness in females more than males?
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
Issues of statistical significance aside, that would be the interpretation as long as you had the level effects as well (say if your model also included a constant as well as experiment and gender coefficients).
