Hello, I am new to mixed effects model and am running a analysis with it.
However, am wondering if the assumptions check for Quadratic mixed models are the same as for linear mixed model?
- Linearity, no outliers, similar spread, normality of residuals
Also if anyone could share about how I could run these checks on SPSS that would be helpful.
Thank you.
Hello Venetta Ho Look, as far as I can try to help you, most of those Assumptions are basically structural. In other words, they are not affected by whether the model is linear or quadratic, or by the transformations you perform on the IVs. So I can say Assumptions you mentioned still hold in a quadratic model.
Hello,Venetta Ho
The relationship between the predictors and the response is linear in terms of the parameters, even though the model includes a quadratic term for the predictor. This means that the model should be correctly specified in terms of including the linear and quadratic terms.
This might be tricky but you need to explain more. Are all the IV's quadratic in nature or some of them only?
In each case, the four basic assumptions should be check... but I will advise you go for GLM which transforms the linear predictor before fitting the data. Knowing that the model will stand valid even if the basic assumptions are not met.
- Badges
- Science topic