Generalized Linear Models

3 units

Generalized linear models are a collection of statistical methods used to analyze categorical and limited dependent variables. In this course, students will learn fundamental concepts and skills to conduct generalized linear models, and know how to apply these techniques to social, behavioral, and health research. The course covers the following topics: the Nelder and Wedderburn framework of generalized linear models and the key concept of link function, maximum likelihood estimator, a review of logistic and probit models, multinomial logit model, ordered logistic regression, Poisson regression, negative binomial regression, quasi-likelihood functions, and model fit/validation. This course is designed to fulfill part of the core quantitative methods requirements for doctoral students at the Brown School. Prior enrollment in BSDC 8005 Applied Linear regression Analysis or a similar course is encouraged.