Given that most communication phenomena involve complex processes across multiple levels of analysis and over time, multilevel modeling (MLM) has unique value for communication scientists by permitting the simultaneous analyses of data collected at more than one level. In particular, [Page 903]the application of MLM enables health communication scholars to stand a better chance of generating valid knowledge relevant to describing and explaining health communication processes because health behaviors and outcomes are determined by both microlevel and macrolevel factors.

MLM, also called hierarchical linear modeling, random coefficient regression, or mixed effects modeling, is a group of statistical techniques to cope with hierarchical data, in which each observation at a lower level is embedded within its higher-level units. To employ MLM, researchers should use one of the ...