Not Quite Normal: Choosing the Best Response Distribution for Modeling Your Response ( 2019-EU-45MP-107 )
Feb 11, 2019 4:32 PM
Level: Intermediate Job Function: Analyst / Scientist / Engineer Clay Barker, JMP Principal Research Statistician Developer, SAS
The Generalized Regression platform in JMP Pro is a flexible and powerful tool for building regression models. While we often talk about the variable selection methods built into the platform, we cannot forget about the ability to specify the distribution of the response. Do we have count data or a skewed response? Or is the normal distribution sufficient? In this talk we will give a brief overview of generalized linear models and then provide some strategies for choosing the best distribution for your response in the Generalized Regression platform.