BIOS 6311: Principles of Modern Biostatistics

ALERT! Space is limited and enrollment requires special permission. Priority is given to graduate students from the Biostatistics, Epidemiology, and Biomedical Informatics programs, for whom the course is required. Most years there are not open spaces after fitting in those three programs. Please email Robert Greevy and Chazlie Miller to apply for the course and to be added to a waitlist.

Course Information

Course Instructor: Robert A. Greevy, Jr, PhD
Lab Instructor:

Class meets Tuesdays and Thursdays 10:30am-12:00pm in 2525 West End Ave, 11th floor, room 11.105 (the large classroom).
and lab meets Tuesdays 3:05pm-4:05pm in the same room as the class.
Both class (311) and lab (311L) are required.

Office Hours: Thursdays 3:15pm-4:15pm in the same room as the class.

Recommended texts:
I will recommend readings from at least three texts: 1. Rice, 2. Rosner, and 3. Verzani. Rice and Rosner are both fairly expensive and are used as optional additional explanatory material. Verzani is reasonably priced and teaches skills in R that are required for, but not taught in, this class. You are not required to buy your own copy of these texts, but you are responsible for learning the content. Use your judgment and please feel free to ask me for guidance specific to your needs. For Biostatistics students, Rice is written at the level of understanding you should get to by the end of the course. For non-Biostatistics students and students with limited prior statistical training, Rosner is a good introduction to methods, basic theory, and traditional views of statistical practice. Rosner has lots of great, medical context driven, problems. The electronic and/or paperback versions may be less expensive.
  1. Mathematical Statistics and Data Analysis by John Rice ( amazon)
  2. Fundamentals of Biostatistics, 7th ed., by Bernard Rosner ( amazon)
  3. Using R for Introductory Statistics, Second Edition by John Verzani ( amazon)

The course requires two statistical programs: R and Stata.
A laptop with R and Stata installed will frequently be required for class and lab.
Methods will be introduced in Stata and their performance studied in R. R can be download for free at The free version of RStudio is recommended as a front end for R, see
Stata can be purchased at a student discount at See to decide which version is best for you. Stata/IC is the best choice for most students.

Class Links

Recommended Resources

R resources

Stata resources

Official course description

6311 & 6311L. Principles of Modern Biostatistics: Principles of Modern Biostatistics is a foundational first course in graduate level statistics designed to develop a richer understanding of one- and two-sample statistical methods and statistical philosophies. It explores the operational characteristics of frequently used statistical methods. Through simulation studies conducted in R and STATA, students will explore questions such as: What are the true coverage rates of commonly used confidence interval methods for proportions? What is the impact of sampling from various non-normal distributions on the true Type I Error rate for hypothesis testing methods? How do various testing methods compare in terms of power in a variety of settings? How do traditional hypothesis testing methods compare and contrast with methods in the Bayesian and Likelihoodist paradigms? This course is intended for graduate students in programs for biostatistics, biomedical informatics, and epidemiology, and by students in other programs who have a strong undergraduate-level background in statistics. Lab required [1]. Prerequisite: Calculus I. Fall. [3]. Greevy.

Unofficial course description

This course assumes familiarity with univariate and bivariate data analysis methods (t-tests, Chi-squared tests, tests based on asymptotic Normality, tests for paired data, etc.) and fundamental statistical concepts (summary measures and graphs, confidence intervals, hypothesis testing, etc.). The primary goal of this course is not to reintroduce these concepts and tools at a more rigorous level, although that is done. The primary goal is to study and understand the performance and behavior of these methods in various settings. When does the 95% confidence interval actually have 95% coverage? Under what settings is the coverage for a method close enough to the claimed 95%. Given a set of assumptions, what does the power curve for a method look like as we vary the effect size? ... as we vary the sample size? ... as we change the population distributions? In addition, the course explores interval and inferential approaches based purely on the likelihood function and on the likelihood with an assumed prior distribution, i.e. the likelihood inference and Bayesian inference paradigms.

Topic revision: r166 - 10 Feb 2020, RobertGreevy

This site is powered by FoswikiCopyright © 2013-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Vanderbilt Biostatistics Wiki? Send feedback