# SCS Reads 2015-2016

### From Wiki1

## Contents |

## Getting started

- We chose to read the book by David Kaplan,
*Bayesian Statistics for the Social Sciences*for the seminar series.- Companion web site with R code and data from the book.
- Zip file containing code and data

- Julia Galef: A visual guide to Bayesian Thinking
- Uses mosaic plots to illustrate the problem with 'base-rate neglect' or, as termed by Kahneman and Tversky, the 'representativeness heuristic'.

- John Kruschke videos on Bayes vs. NHST
- Three short videos comparing the Bayesian approach to NHST and frequentist approaches from talks and conference presentations.

- Some additional presentations and overviews of Bayesian Statistics:

## Week 1 (Oct. 9)

We are reading the very short Chapter 1 from Kaplan

- An interactive demonstration of the Monty Hall problem addressed in Chapter 1 of Kaplan

- Brief_history_of_Bayes1-2x2.pdf Michael Friendly's slides on early history of Bayes

- BayesBalls.R Animation of Bayes billiard thought experiment

- John Fox's lecture notes on probability: Fox-Lecture-10-notes.pdf and Fox-Lecture-12-notes.pdf

- John Fox's Appendices to Applied Regression also contains one on probability theory.

## Week 2 (Oct. 23)

- Kaplan, Ch. 2 Kaplan_BSSS_Ch.2.pdf

- To complement Kaplan's presentation: Jeffreys and uniform priors for the probability of success in a Bernoulli trial Priors-Jeffreys-uniform.pdf (corresponding R code: Priors-Jeffreys-uniform.R)

- To see the shape of the beta distribution: beta-plot.pdf (corresponding R code: beta-plot.R)

- Bernoulli distribution example--- how to specify a beta(a, b) prior as prior mean and effective sample size: bern-beta-ex.pdf (corresponding R code: bern-beta-ex.R)
- This uses some R functions from
*Doing Bayesian Data Analysis*, [DBDA software page]

- This uses some R functions from

## Week 3 (Nov. 6)

- Kaplan, Ch 3. Another ho-hum chapter, giving results for priors and posteriors for a variety of standard distributions without much

insight. Would make a reasonable Wikipedia page.

- Demonstration of Poisson distribution with Gamma prior (Fig 3.4) poisson.pdf (corresponding R code: poisson.R)

## Week 4 (Nov. 20)

- Kruschke's Chapter on Markov Chain Monte Carlo: kruschke_ch7_mcmc.pdf

- A nice example of the Metropolis-Hasting MCMC algorithm in R: https://theoreticalecology.wordpress.com/2010/09/17/metropolis-hastings-mcmc-in-r/
- R code for this example: MCMC-demo.R

- An R function to illustrate a bivariate Gibbs Sampler for normal distribution: biNormGibbs.R
- Gibbs sampler demo: BiNorm-demo.R using this function

- Jamie Kim suggested this YouTube video: MCMC & Metropolis-Hastings sampling

## Week 5 (Dec. 4)

- Reading: Kaplan Chapter 5,
*Bayesian Hypothesis Testing*

- JAGS analysis of simple linear regression jags-ex1.R

- Output: jags-ex1.pdf

- BayesFactor examples bayes-factor-ex.R

- Output: bayes-factor-ex.pdf

## Week 6 (Jan. 15)

- Reading: Kaplan Chapter 6,
*Bayesian Linear and Generalized Linear Models*

- PISA2009 data: PISA2009.csv

## Week 7 (Jan. 29)

- Reading: Kaplan Chapter 7,
*Missing data from a Bayesian perspective*

## Week 8 (Feb. 12)

- Reading: Kaplan Chapter 8,
*Bayesian multilevel modeling*

## Week 9 (March 4)

- Reading: Kaplan Chapter 9,
*Bayesian modeling for continuous and categorical latent variables*