Perhaps most importantly, we are also able to do cross-level interactions so that we can explain how Level 1 predictors affect our dependent variable according to different contexts (Level 2 predictors).

The information about individual and group characteristics is retained and separate estimates are produced for both.

Additionally, we can look at how Level 1 predictors interact with each other and how Level 2 predictors interact with each other.

If you violate it, you get incorrect estimations of the standard errors.

Multilevel modeling is an extension of OLS.

relationship between the independent and dependent variable is linear.

The assumption of independence means that cases in our data should be independent of one another,

independent variables are measured without error.

the sample used in the analysis is representative of its population.