My team has a multivariate normally distributed data that came from a snowball sample. We are wondering if we can apply a mixed-effects regression (MER) to analyse this data
I would tend to agree with Mark Hanke . Your non-random sampling approach does not mean you cannot test the assumption that your data is normally distributed.