I have an independent variable (a bias) that is technically a constant. I want to measure the effects of the IV on the DV. How do I do this? I'm using SPSS.
Let's say, participants produced A or B answers (DV) but according to the bias they should have produced only A answers. There are other factors involved, including other biases (binary) for which there was no significant main effect (GEE analyses). So now I need to know if/how much the bias (the one that's a constant) is responsible for the results, but because it's read as a constant it's assumed to be redundant. This means I can't run GEEs testing interactions between the two 'bias factors' without getting "empty coefficients". Help?