Linear mixed models are the study of fixed and random effects. Your fixed effects are usually risk factors which you consider to be predictors of the dependent variable. Your random effects are data attributes which promote variability amongst the subjects in your study. An example would be cases where you have a repeated measures study where not every patient/subject has the same number of cases. So you may have visits 1,  2, and 3 for some subjects and not others. Random behavior within and among subjects, such as visit number is a scenario for a linear mixed effects model. Assuming your dependent variable is a scaled measurement and you have random behavior or effects. Know your data.

-Moore to Follow Amy