Bayesian Data Analysis Materials 2017-06-22T18:01:12+00:00

Bayesian Data Analysis Materials

This page contains links to materials and resources associated with Bayesian Data Analysis course.

Please click on the link below:

Materials and lectures

1.A. Before arriving at the workshop, be sure to install software on your notebook computer as explained at this website:
https://sites.google.com/site/doingbayesiandataanalysis/software-installation
and be sure to get the latest programs that accompany the book as described on that page, also available at this link:
https://sites.google.com/site/doingbayesiandataanalysis/software-installation/DBDA2Eprograms.zip?attredirects=0&d=1
The various software components are designed to play nicely together only with synchronized recent versions, therefore if you have old versions of any of the software, please uninstall the old versions are install the latest versions.
1.B. The Stan software will only be introduced briefly if there is time and interest; therefore installing Stan is optional.

2. The book is optional but recommended. If you are interested in purchasing a copy, notice that there is presently a 25% discount, and the link also has a code for a 30% discount, on both print and electronic versions, at the book’s web site:
https://sites.google.com/site/doingbayesiandataanalysis/purchase

3. Presentation slides are now available for download. The PDF has 949 pages, so save trees and don’t print it, but you may find it useful to have the slides available on your notebook computer and to follow along during the class. The 30 MB file is available for download here:
https://www.slashtmp.iu.edu/files/download?FILE=kruschke%2F56502usrjY9
The decryption password is: Bayesian

4. Optional: Bring along some of your own data to try in the analysis programs, in a format specified below.
4.A. Two groups of metric data. The file format should be comma-separated value (.csv). It should have two columns. All columns should have a column name on the first line. One column should specify the metric data values, and the other column should specify the corresponding group membership of each datum (i.e., group 1 or group 2). In other words, one <y,g> pair per row. There should be no missing data. Make it a reasonably data set, not one with thousands of rows.
4.B. Simple linear regression. The file format should be comma-separated value (.csv). It should have two columns. All columns should have a column name on the first line. One column should specify the metric Y values, and the other column should specify the corresponding metric X values. In other words, one <x,y> pair per row. There should be no missing data. Make it a reasonably data set, not one with thousands of rows.
4.C. Hierarchical regression, a.k.a. panel data. The file format should be comma-separated value (.csv). It should have three columns. All columns should have a column name on the first line. One column should specify the metric Y values, a second column should specify the corresponding metric X values, and a third column should specify the panel (subject, unit) from which the x,y data came (an integer or name). In other words, one <x,y,p> triplet per row.  There should be no missing data. Make it a reasonably data set, not one with thousands of rows.