Search | Directories | Reference Tools
UW Home > Discover UW > Student Guide > Course Catalog 

Instructor Class Description

Time Schedule:

Beth A. Mueller
EPI 514
Seattle Campus

Application of Epidemiologic Methods

Practical experience in analysis of data. Students analyze data sets currently on file using contemporary epidemiologic methods as taught in EPI 512 and EPI 513. Prerequisite: EPI 510 or experience in statistical programming; EPI 512, EPI 513, and epidemiology major. Offered: Sp.

Class description

The purpose is to give graduate epidemiology students the opportunity to gain “hands-on” experience analyzing data to answer a research question. Methods and theoretical issues introduced in EPI 512-513 will be briefly recapped; however, the focus is on practical analysis issues with real data, rather than further theoretical discussion. Students work in small groups (2-3) on separate research topics, with each group using a different dataset that is provided.

Student learning goals

Students will gain experience in conducting a research project from start to finish, by carrying out necessary development, writing, and analysis tasks related to their projects. At the end of this course, students should be able to:

1. Search scientific literature to identify relevant reports of epidemiologic and other studies related to a research topic, and biologic or other rationale to test a research hypothesis.

2. Formulate focused research hypotheses that can be answered using available class data.

3. Calculate statistical power/sample size for a case-control or cohort study. Identify the most appropriate statistical calculation to present (minimum detectable relative risk/odds ratio, available statistical power, etc.) and formulate an appropriate table presentation and grant proposal text.

4. Compare the advantages and disadvantages of different study designs for a research project in order to identify, justify, and describe an appropriate study design based on a research question, available data, and statistical power and sample size.

5. Identify and list potential confounders and effect modifiers of a disease/outcome-exposure association, while considering issues of data quality and availability, prior published or clinical knowledge, and sample size/power.

6. Develop a written proposal, using a standard NIH format with specific aims, background and significance, study design, sample size calculations, and limitations.

7. Conduct computer programming necessary to prepare a data file for analyses including data cleaning runs (checking ranges, missing values, etc.), and basic file management (categorizing variables, etc.) necessary for stratified analyses.

8. Perform exploratory data programming necessary to present basic descriptive tables comparing categorical variables between 2 or more study groups.

9. Analyze data to evaluate associations using Mantel-Haenszel methods and stratified analyses to estimate odds ratios and relative risks.

10. Estimate risks estimates and/or rates for a disease/outcome-exposure association, evaluate confounding and effect modification, and identify the most appropriately adjusted or stratified presentation of risk estimates to answer a research question.

11. Identify and describe the limitations of epidemiological analyses (e.g. related to data quality and availability, statistical power, etc.) for answering a research question.

12. Organize and display research findings in standard table formats for epidemiologic journals.

13. Present final study results in the format of a scientific paper for publication including an abstract, introduction, methods, results, and discussion of findings.

14. Integrate and correctly format scientific references in a scientific paper for publication.

15. Develop and present a scientific poster or talk (a scientific abstract) of study results.

General method of instruction

There will be a faculty lecture each Tuesday. The TAs will present material in the laboratory sessions on Thursdays pertaining to the practical aspects of data analysis and statistical computing.

Recommended preparation

Students should come into the class with some knowledge of computing so that they do not spend an inordinate amount of time trying to learn programming concepts (EPI 510 or equivalent experience).

Class assignments and grading

The primary emphasis is on the group project. Students will work in small groups on separate research topics, with each group using a different dataset. Each group will have a primary faculty preceptor (assigned by the instructors) with whom they will work most closely. Each group will develop a focused research question, conduct a literature search to ascertain the biologic or other rationale to test a certain hypothesis, select an appropriate study design based on the research question and issues related to power and sample size, write a proposal, using a standard, NIH proposal format, to present the research question, conduct preliminary data cleaning runs, perform exploratory descriptive data analysis, conduct a stratified analysis, interpret results, estimate crude and adjusted risks and/or rates, assess confounding, effect modification, trends, etc., write up final results in the format of a scientific paper for publication, and present results in either a poster or talk format at the class conference.

Most of the graded course work is related to the class research project, and is turned in by the project team. The first stratified analysis homework is to be turned in by each individual. Weights assigned to each assignment are as follows: Stratified analysis assignment which is turned in individually, not by groups (10%), project proposal (15%), stratified analysis of project (15%), final presentation/poster (25%), participation (lab, class interaction, etc., 10%), final paper (25%).

The information above is intended to be helpful in choosing courses. Because the instructor may further develop his/her plans for this course, its characteristics are subject to change without notice. In most cases, the official course syllabus will be distributed on the first day of class.
Last Update by Angelica M. Buck
Date: 05/09/2013