Presenter: Brian Fogarty
This workshop focuses on techniques when your outcome variable is not continuous or normally distributed. Such outcome variables are considerably more common in social science data than normally distributed outcome variables. We will briefly explore four different types of outcome variables: binary, ordered, unordered, and count variables. For point of reference, the material in this course is referred to using several different titles including generalized linear models, categorical data analysis, and limited and categorical dependent variable regression. The topics covered include:
- Binary outcome models (logit/probit)
- Ordered outcome models (ordered logit/probit)
- Unordered outcome models (multinomial logit/probit)
- Count models (Poisson, negative binomial)
- Statistical significance and regression coefficient interpretation
- Visualizations of regression coefficients
The workshop is designed for individuals with experience using generalized linear regression in other statistical software (e.g., Stata, SPSS) and want to learn how to run these models in R. Further, it is assumed that individuals have a basic understanding of R. The workshop is open to undergraduate students, graduate students, faculty, and staff. Although the workshop uses social science data, people from other disciplines are welcome to attend.
This workshop will be offered on Zoom. There is a limit of 15 participants for this workshop. Register Now!
Registration must be completed by Monday, March 14th.
More details about the workshop can be found here.
