This course covers a novel approach to dealing with systematic error in epidemiologic research called quantitative bias analysis, QBA. QBA allows users to make corrections to measures of association for sources of bias by making assumptions typically using validation data about the nature of the bias to create intervals that account for total study error. The course will cover three types of bias analysis: simple, multidimensional and probabilistic. Exercises in Excel and SAS will allow students to practice the methods, adapt them to problems they face and present the results clearly.