Multilevel Modeling Workshop

Shevaun Neupert, Ph.D.

The Multilevel Modeling workshop will briefly introduce the logic of the method (including when and why to use it), but most of the time will be spent on application where participants will analyze and interpret their own data. Multilevel models are used (or at least they should be!) when individuals are nested within larger groups such as communities, companies, or classrooms, and/or when individuals are measured repeatedly over time.

Analyses will be conducted using SAS, but templates of necessary code will be provided to participants so that only minimal alterations will be required. Preparation of datasets will be completed in consultation with Shevaun Neupert prior to the beginning of the workshop so that participants can analyze their data on the first day of the workshop. By the end of the workshop participants will be able to analyze their data with multilevel models relevant to their research questions and also interpret the results so they are ready for publication. Specific topics will include:

• Partitioning variance and calculating the intraclass correlation
• Grand-mean and group-mean centering
• 5 basic multilevel models that provide the structure for all models
• Graphing cross-level interactions
• Calculating effect size

Back to Main Page