Speaker: Malte Schierholz MZES, Mannheim, Dec 14 2016
Besides the frequentist approach to statistical inference, which was dominant in science in the 20th century, another school exists: Bayesian Statistics. With modern computational techniques, Bayesian data analysis has a proven track-record and established itself as an alternative to frequentist procedures. Sometimes, Bayesian techniques can be applied to complex scientific questions where no frequentist solution exists.
This talk gives an introduction to Bayesian statistics. While it is not possible to avoid central mathematical formulas and derivations, I concentrate on concepts, intuitive motivations, and interpretations that underlie the Bayesian view. Critical model assumptions are also discussed. Participants will learn when to mistrust a Bayesian analysis and in which situations it may provide new insights.
Available here
http://andrewgelman.com/2016/12/13/bayesian-statistics-whats/ https://www.crcpress.com/Statistical-Rethinking-A-Bayesian-Course-with-Examples-in-R-and-Stan/McElreath/p/book/9781482253443 http://www.stat.columbia.edu/~gelman/arm/ http://www.stat.columbia.edu/~gelman/book/