Handling Non-independence with Mixed Models

In analyzing data from quantitative studies, researchers often run into a common statistical challenge: The observations aren’t independent from one another. For example, if an ecology study is conducted across multiple geographic locations, observations may not be independent due to uncontrolled environmental factors that differed across locations. Or if a survey study involves repeatedly surveying the same people over several months, any individual participant’s responses are likely to be correlated across time. Conducting standard regression on non-independent data can lead to inaccurate inferences, but thankfully, there are statistical techniques that allow you to account for non-independence in your data.

This webinar, intended for academic researchers, teaches you concepts and techniques for handling non-independence using linear mixed effects models, including how to implement these techniques in the JMP Student Edition, a no-cost, full-featured version of JMP exclusively for academic use. Topics include:

• Mixed models for non-independence among grouped or clustered units of observations

• Mixed models for non-independence in repeated measures taken across time or space

• Generalized linear mixed models for non-independence and non-normal response data

Get JMP software free for academic use at jmp.com/student