Finally we come to a close to discuss the assumptions that our random intercept and slope models rely on.

So far we have simply believed that Stata and R give us unbiased standard errors that allow for accurate statistical inference. In this session we learn that it is not that simple.

In this session we will extent the random intercept model in such ways that slopes of micro-level variables may also vary across clusters.

In this session we will use our new knowledge on how to model random slopes and cross-level interaction to replicate a recent study on symbolic and realistic immigration-threat perceptions.

In this session we will use our new multilevel modeling skills to replicate a recent study on xenophobia.