 |
 Categorical data analysis is
an important tool in many areas, particularly biological and
health sciences. This type of analysis is focused on outcomes
that either cannot or should not be studied using a continuous
model. The most common type of categorical analysis is with a
binary yes/no outcome such as presence or absence of disease or
success or failure of a process. Since this type of outcome is
so common, we will spend a large proportion of the course
working with this sort of data. We will learn to analyze binary
outcomes like this in detail using univariate techniques and
logistic regression. In particular, we will focus on
interpreting and reporting binary outcomes and their predictors
in a fashion that makes our results understandable to the end
user. In addition, we will work with techniques for modeling
multi-level outcomes and survival data, which are also common in
today’s world. We will discuss how to make decisions about using
various categorical models for both predictors and outcomes. At
the end of the course, the students should be able to conduct
and report a complete analysis of several types of categorical
outcomes.
|
