What might be the most appropiate approach to modeling an ordinal response variable in a DoE ?
Hi Community, I have a DoE with several response variables. One of those response variables is an ordinal one having levels: -3, -2, -1, 0, 1, 2 and 3. My questions is what might be the best or more pertinent way to model this response variable ? is it to model it as an ordinal variable using logistic regression ? or considering the variable as a continuous one and using traditional Ordinal least ...