introductory statistical methods courses, students are taught a set of techniques that most often require data to be models using a normal distribution. Linear models built in this manner have a number of useful properties. However, often particular types of data, such as categorical or count data, cannot be adequately modeled by a normal distribution. Often the solution to this problem is to apply some arbitrary transformation to the data. Often, however, this is also inadequate. A better solution to the problem is to abandon the use of a normal distribution and use a statistical distribution that better reflects the properties of the data. A generalized linear model is based on the concept that to test a scientific hypothesis using a statistical distribution other than a normal distribution, it is necessary to take a function of the parameters of the statistical distribution being used to relate to the linear hypothesis being tested. This workshop will review some basic generalized linear models and give examples of how to use the R statistical software system to analyze them.