EPID 652: Advanced Epidemiologic Methods and Data Analysis
Course Prerequisites:
EPID 651 and BIOS 554, or equivalent
Course Description:
This course focuses on development of analytical strategies for data analysis guided by
epidemiologic principles. Specific statistical modeling will be tailored for analysis of data from cross-sectional,
case-control and cohort studies with emphases on effect estimates, causal inference, controlling for
confounding, and assessment of interaction and mediation effects. Course topics cover applied multivariate
analysis including logistic regression, log-binomial regression, Cox proportional hazards model, Poisson
regression, multilevel and path analytical model.
Course Goal:
The goal of this class is to provide epidemiologic principles and statistical techniques that are
commonly used in advanced epidemiologic research. After the course, students will be of confident in the
application of advanced modeling techniques in analysis of epidemiologic data and interpretation of findings.
Objectives:
By the end of the course, the student will
- Develop analytic strategies for various study designs, guided by the principles of epidemiology.
-
Understand the rationale and assumptions underlying the major statistical techniques used to analyze
data from epidemiological studies.
- Describe strengths, limitations, and issues pertinent to the proper application of these techniques.
-
Explain how interactions/effect modification, confounders and dose-response relationships among
variables are examined.
-
Utilize appropriate methods of analysis to account for interaction/effect modification, confounding, and
mediation effects.
-
Compute and interpret odds ratios, prevalence ratio, relative risks, rate ratio, hazard ratio, and their
confidence intervals.
-
Demonstrate competency in using SAS to perform data analysis for cross-sectional, case-control, and
cohort studies.
Required Textbooks:
-
Rothman KJ, Greenland S, Lash TL. Modern Epidemiology, 3 rd edition. Lippincott Williams &
Wilkins, Philadelphia, PA, 2008.
-
Logistic Regression: A self-learning text (3rd Edition) D.G. Kleinbaum and M. Klein. Springer,
2010
- Handouts will be on canvas or distributed in class if necessary
Supplemental text:
-
Szklo M and Nieto FJ. Epidemiology: Beyond the Basics, 3 rd edition. Jones & Bartlett Learning,
Burlington, MA, 2014.
-
Woodward M. Epidemiology: Study Design and Data Analysis, 2 nd edition, Chapman & Hall/CRC,
Boca Raton, FL, 2005.
Course Format
Course sessions will entail a combination of lectures, discussions, problem sets, and in-class activities in the lab
sessions. Based on by our teaching philosophy and experience, we believe active student participation is
essential in success.
There are three parts in the 3-credit course, (1) overview of epidemiologic designs, e.g., cross-sectional, case-
control, and cohort studies; (2) application of specific statistical methods in a specific study design; (3) use of
SAS statistical software to perform data analysis. The first two items will be covered during classroom lectures
and the third part will involve SAS programming.
Course Evaluation
Final grades will be determined by scores of class participation, Exam 1, homework assignments, and Exam 2.
Grading
- 10% – Class participation (all absences will be discounted equally)
-
30% – Homework assignments. Each assignment will be worth 10 points; 3 of these will be given for
completing the assignment on time, and 7 points will be based on correctness and completeness of
answers.
- 30% – Exam 1
- 30% – Exam 2