my statistical professor has just suggested me to use mixed model for my study and unfortunately I am not familiar with this model.
Before having discussion with him, I have analyzed the data with ANOVA, paired t test and multiple linear regression.
My questions are:
1. When do we usually use a mixed model?
2. Is mixed model one type of multivariate analysis?
3. Should I reanalyze group differences (previously I used ANOVA) and within group differences (from first measure to final measure, previously I use Paired T test) and how can I run such kind of tests for a mixed model?
4. For running mixed model, is that true that the distribution of the data should be normal? How is when data distribution not normal?
5. If I have weight measurement before intervention and after intervention, should I first change the data from wide data format to long data format?
I am so sorry that I write the questions in english, but please feel free to answer it in english or german
.I thank you very much before.
LG
