PHS 706 - Paradigms for Statistical Inference

Frequentist, Bayesian, and Likelihood-based paradigms provide alternative paradigms for drawing conclusions from data. Each addresses a different question of interest. This course compares the questions asked and properties of these three paradigms. A portion of the course will be devoted to learning Bayesian analysis. Topics will include: measures of statistical evidence; single versus multiple hypotheses; simple versus composite hypotheses; impact of accumulating data upon conclusions. Students will learn arguments for and against each paradigm and will be left to draw their own conclusions on how to approach analysis. The course is intended to bridge the mathematics of inference with practical application.