This website is dedicated to the support of workshops taught by Larry Madden and colleagues. Most of the workshops are taught at the annual meetings of the American Phytopathological Society (APS) and of other scientific societies. Please go to the pages for Instructional Material and Downloads to obtain files for the upcoming and past workshops.
Sorry, but this workshop is over. See you next time!
Mixed Models for Data Analysis in Plant Pathology
L.V. Madden (Ohio State Univ.)
A.B. Kriss (Syngenta)
At the 2019 Annual Meeting of
The American Phytopathological Society
Huntington Convention Center of Cleveland
Saturday August 3rd
Room 16 from 1pm – 5pm
Participation in this workshop requires prior registration
It is now standard practice to use linear mixed models (LMMs) and generalized linear mixed models (GLMMs) to analyze data from designed experiments. Mixed models formally handle experiments with both fixed (e.g., fungicide treatment) and random (e.g., block, location) effects, account for correlations of the observations (inherent to the design), and properly estimate test statistics and standard errors. The transition from ANOVA to mixed models requires a major change in perspective and understanding, yet few researchers are formally taught the methodology. Registrants will learn the concepts underlying mixed models for normal and non-normal data, with an emphasis on analyzing experiments that are repeated in space or time. Researchers often are concerned about pooling or not pooling data from multiple runs of an experiment, and mixed models provide a formal way to make this decision. We will also place a special emphasis on the calculation and interpretation of BLUPs (Best Linear Unbiased Predictors) that are fundamental for characterizing random effects. SAS procedures will be used for all analyses. This workshop will follow the tradition of the Epidemiology Committee’s on “bringing statistical analysis to the masses”.
- Registrants need to bring a laptop with SAS 9.3 (or later) installed. The free SAS University Edition is also acceptable.
Location of the Workshop
The next workshop is yet to be determined.