# R for Starters, v0.57741

The book,

`R`

for Starters, is divided into four parts. The goal of the book is to cover the theory, as well as the implementation of the theory. To that end, it entwines the equations with the code in an effort to allow people to *do*statistics, rather than just read about it.
The first two parts cover the typical topics in an introductory statistics course. These include the relationship of statistics to science, and the theory behind statistical testing. The first part includes an introduction to

`R`

and to statistics. It covers installing and using `R`

, the scientific method, and simple descriptive statistics.
The second part explores inferential statistics, focusing on the assumptions needed to create the statistical tests. This includes tests with discrete independent variables, both discrete vs. continuous and discrete vs. discrete analysis. For those with a background in elementary statistics, use this as both a review of some statistical ideas and as an introduction to using

`R`

. It will allow an easy entry into scripting while not having to learn new statistics. Here, we will also learn how to produce some powerful graphics that can be easily saved and used in your research, both as graphical tests and as production-quality graphs.
Part Three covers the classical linear model, as well as the ordinary least squares method for estimating the parameters. Emphasis on assumptions continues here, as well as what to do when the assumptions are violated by the data and the model.

The fourth part deals with a systematized improvement to the Classical Linear Model: the Generalized Linear Models (GLMs). This type of model is quite prevalent in the research, as it allows one to easily handle dependent variables that are non-continuous or are bounded. The Classical Linear Model (CLM) makes several assumptions about the dependent variable being analyzed, the most limiting of which are the assumptions that the dependent variable is continuous and unbounded. GLMs relax those two assumptions, thus allowing us to model count data, binary data, proportion data, and survival data.