# Course Description: BIOS601: Epidemiology, Introduction & Statistical models: Fall 2019

## [updated August 06, 2019]

Instructor Dr. James Hanley
Co-ordinates tel: (514) 398-6270
e-mail: james.hanley@mcgill.ca
Web-page: http://www.biostat.mcgill.ca/hanley
Overview The course covers the elementary statistical principles and methods used in data-analysis in scientific and population-health research, and the theoretical (mathematical-statistics) foundations for these methods. The applications illustrated will be drawn mainly from epidemiologic and clinical research; portions of the classes and assignments will involve mathematical statistics; others will require some statistical computing.
Target This course is aimed at MSc and PhD students in the department's biostatistics program, i.e., students with undergraduate training in mathematical statistics who are interested in the application of statistical methods to epidemiologic, biomedical and other biological research. However the skills and understanding acquired in this course are also applicable to other fields of research that use quantitative methods. Statistically-prepared students from other departments or universities are especially welcome.
Course Website http://www.biostat.mcgill.ca/hanley/bios601/index.html
Topics 1a: Epidemiologic concepts, terminology and measures.
1b: Surveys, sampling, and measurement
1c: Statistical models, inference and planning for such (1-sample) studies.

2a: Comparative studies [experimental, non-experimental, quasi-experimental]
-- in general
-- specifically in epidemiology (with associated terminology and measures).
2b: Statistical models, inference and planning for comparative studies.

3: Study designs & statistical methods for reducing bias & increasing precision.
Approach

## It is assumed that students will have already encountered the basic statistical models (Gaussian, Binomial and Poisson random variables) and basic inferential procedures. The distinguishing features will be the • use of likelihood as a unifying theme • integration of study design, data analysis models, and precison/power/sample size • coverage of both larger- and smaller-sample situations • focus on a minimalist approach, and on concepts, so that one can more easily go from a familiar to an unfamiliar statistical problem • reference to (generic) classic texts by Cox (Planning of Experiments), Cochran (Observational Studies), Campbell and Stanley (Quasi-experimental Designs), and the specific one by Clayton and Hill (Statistical Models for Epidemiology) which uses likelihood as a unifying theme.

When Mondays and Wednesdays 13:30-15:30.     first/last class: Sept 4/Dec 3.
Where Room 24, first Floor, Purvis Hall, 1020 Pine Ave. West [corner Pine]
Prerequisites Undergraduate course in mathematical statistics at level of MATH 324, or permission of instructor.
No. of Credits 4
Assessment To be discussed at first class: suggested:- assignments [xx%], participation [xx%], exams[xx%]