ABSTRACT
This chapter explains when to use Multilevel Modeling (MLM), why nested data require MLM, how common models work, how to interpret output from MLM software packages, and how MLM has been applied in planning studies. The origins of multilevel modeling trace back to the nineteenth century, when sociologists began researching contextual effects on individual behavior. MLMs sometimes represent data structures where each group belongs to an even higher unit of organization: one level of data is nested within another level of data. By the 1980s, MLM techniques were used in a variety of social science fields, but perhaps no application became as iconic as education research. MLM enables an analyst to explain the variance at each level using variables specific to that level. MLM software generates an intercept and two types of estimates for regression coefficients.
