While analyzing the effect of level 2 variables on level one outcome variable, only level one related control variables are required to add or also the level 2 control variables are also required to include as controls?
If there are level 2 variables you want to control for then you should include them into your model. When using R (for example lmer function) you have to check wether your level 2 control variables should be included in the random intercept part only or also in the random slope part(s).
Well, I think there is no difference between variables you want or have to control for on level 1 and control variables on level 2. An (maybe very artificial) example: If you are analyzing the effect of the age of a therapist (level 2 variable) on the symptoms of the patient after treatment (level 1 outcome), there are of course level 1- and level 2 variables you could control for. On level 1, you should/could control for the symptoms of the patient before treatment, on level 2 you could control for the years of training of the therapist/ how long he is working as a therapist.
Concerning my second point, using my example: The next question is wether there is a predictor variable on level 1. If not, the years of training can only have an impact on the intercept of the level 1 regression. If there is a predictor variable on level 1, you have to reflect if the years of training have a potential impact only on the intercept of the regression or also on the regression slope.
hello brian,
how to run control variable in spss and which part of mixed model analysis that we have to put our control variables? thank you Brian Schwartz
You just have to put your control variables as additional precitor variables in the model
- Badges
- Science topic
- Similar topics
- Biological Science
- Physiology
- Physiological Processes
- Stress