Amanda I Phipps
Introduces advanced epidemiologic methods, including causal modeling, inverse probability weighting, propensity scores, sensitivity analysis, imputation for missing data, approaches to multiple comparisons, Bayesian adjustment of risk estimates, recursive portioning, and modeling for prediction. Prerequisite: EPI 512; EPI 513; EPI 514; EPI 536/BIOST 536. Offered: jointly with BIOST 519; Sp.
The primary objective of this course is to deepen studentsí knowledge of epidemiological and biostatistical principles by focusing on emerging or less traditional methodological approaches to handling common problems in epidemiologic research. This course will also explore how these advanced methods relate to the philosophical and theoretical underpinning of more traditional epidemiological methods.
Student learning goals
1) Gain proficiency in critically reading and interpreting epidemiologic studies that apply advanced methods for analysis and interpretation
2) Develop a recognition of how advanced methods relate to more traditional epidemiologic methods
3) Demonstrate an ability to identify appropriate situations for the application of various advanced epidemiologic methods
4) Demonstrate the ability to self-direct learning about a specific analytical problem and critically evaluate alternative statistical methods to apply to the situation
5) Effectively converse with epidemiologist and biostatistician colleagues about application of advanced epidemiologic methods, in a manner necessary for successful collaboration
General method of instruction
The class will be structured with short lectures complemented by in-depth class discussions.
Students should have taken EPI 512/513, EPI 514, and EPI 536/BIOST 536.
Class assignments and grading
Students will be expected to participate in in-depth in-class discussions relating to assigned readings, and to answer questions related to selected assigned readings in the form of brief written assignments. Additional homework assignments will provide students with an opportunity to apply analytic methods reviewed in class using hypothetical data. For the final project, students will be expected to present, in groups, a brief lecture and in-class discussion relating to a method not previously presented in class.
Participation 20% - Journal assignments 20% - Homework assignments 30% - Final project 30%