Hello, I have a sample of around 300 pairs of twins. I would like to test the main effects and interaction between the children's behavior and their parents' behavior toward them, on both a continuous variable and a categorical variable. Parents answered about their behavior toward each twin separately, so there is no higher hierarchy variable which is shared between the twins, except for their family number.
Should I be using TYPE=COMPLEX or TYPE=TWO LEVEL? I tried both methods, and they yield very different results.