Y655: Longitudinal Data Analysis
This course will introduce students to the theory and practice of longitudinal data analysis. We will examine a number of approaches for modeling change (in continuous outcomes) and event occurrence (in categorical or ordinal outcomes), including a careful treatment of the metric of time. Examples for the course will come primarily from the field of education; however, the methods are presented in general and examples are easily extended to many fields. Students will learn to develop, implement, interpret, and report research involving longitudinal data analyses. Further, students are expected to gain proficiency in SAS as it pertains to longitudinal data analysis.
Prerequisites: 2 semesters of graduate coursework in statistics.
Algebra Required?: For notation.
Calculus Required?: For concepts and to a limited degree, for derivations.
Day(s) per week offered: 2 days per week, usually Tue/Thu. Yes, there is a lab as required.
Recommended follow-up classes: Depends on research interests.
Substantive Orientation: Social sciences; however, applications to the natural/physical sciences abound.
Statistical Orientation: Frequentist.
Books used: Singer & Willett. (2003). Applied Longitudinal Data Analysis. Similar, although a bit more technical is Hedeker & Gibbons. (2006). Longitudinal Data Analysis.
Applied/Theoretical: applied with explanations (not many derivations) of statistical theory.
Software Used: SAS
How the software is used: Computation, some programming, and data analysis.
Problem Sets: Yes, with a mix of theoretical/conceptual/analytic problems and applied, data analysis.
Data Analysis: Yes, for the homework assignments and the final project.
Presentations: Yes, a research paper is required for the final project. Can include methodological or applied perspectives.
Exams: None at present; however, that could change.
Keywords: Longitudinal analysis; multilevel model; survival models; data analysis.