- Lecturer
- John M. Miyamoto Yamagishi Kimihiko

- Place
- Intensive

- Credits
- Lecture1 Exercise0 Experiment0

- Code
- 65063

- Syllabus updated
- 2012/6/3

- Lecture notes updated
- 2012/3/26

- Semester
- Spring Semester

Recent developments in Bayesian statistical methodology make it possible to change the statistical methodology of a research-oriented psychology department from one based on classical statistics to one based on Bayesian statistical methods. My talk will review the theoretical and computational developments that now make it possible for non-mathematical research psychologists to use Bayesian statistics in teaching and research.

Recent developments in Bayesian statistical methodology make it possible to change the statistical methodology of a research-oriented psychology department from one based on classical statistics to one based on Bayesian statistical methods. My talk will review the theoretical and computational developments that now make it possible for non-mathematical research psychologists to use Bayesian statistics in teaching and research. I will illustrate how high quality, free, open-source software (R, OpenBUGS and JAGS) make it possible for anyone with a standard psychology background in statistics can start using Bayesian computations in their analysis of data. I will argue that the theory and computational methods of Bayesian statistics have matured to the point where non-mathematical psychologists can use Bayesian ideas in their thinking, and in practical data analysis. I have three general arguments for why the shift towards Bayesian statistical methods will benefit psychology: (i) strictly on theoretical grounds, Bayesian statistical methods are better than classical statistical methods; (ii) Bayesian ideas will help psychologists understand the behavior of organisms in uncertain environments (this is not a new idea to students of judgment and decision making nor, for that matter, of animal cognition); (iii) Bayesian models play a major role in neuroscience and a Bayesian emphasis in the standard statistics curriculum will help students understand these models. These advantages should accrue to psychologists in every specialty. By June 7, 2012 I will upload a webpage for the workshop to https://faculty.washington.edu/jmiyamot/bayes/bstats.htm. This webpage will show how to install the R, OpenBUGS and JAGS software on a Windows 7 computer, or the R and JAGS software on a Mac (OpenBUGS does not run under Mac OS, but JAGS is functionally very similar to OpenBUGS; it is even possibly better than OpenBUGS). It will also point the reader to background for the workshop.

ALL SESSIONS MEET IN WEST 9 BUILDING

June 13, 16:50-18:20, Room 202

Yamagishi introduces the psychological argument in the 20th Century. The gist of the argument claims that humans do not intuitively follow Bayes' rule in their data evaluation. Instead, they readily overlook relevant pieces of information in Bayesian thinking, such as sample size and the population proportion of the event of interest (known as base-rate neglect). These views are regarded as outdated today, yet they are worth knowing from the historical perspective.

June 20, 10:45-12:15, Room 706

Miyamoto will show how to use R, OpenBUGS and JAGS to analyze standard psychological research problems from a Bayesian perspective. He will argue that the Bayesian approach has two major advantages. First, it provides straightforward, meaningful answer to the kinds of questions that researchers ask when attempting to interpret research data. Second, it provides an excellent conceptual framework for thinking about decision processes which are so prevalent in psychological research. Although the focus of the talk will be on understanding how to compute Bayesian statistical analyses, it will hopefully be clear that learning to work in this computational framework also helps one think about theoretical questions in psychological research.

June 20, 13:20-14:50, Room 706

This will be a hands-on workshop in elementary Bayesian statistics. Miyamoto will distribute illustrative code for performing analyses in R, OpenBUGS and JAGS. The code and any datasets will be available at https://faculty.washington.edu/jmiyamot/bayes/bstats.htm after June 7. Explanations will also be available at this webpage for installing R, OpenBUGS and JAGS on a Windows 7 computer, or R and JAGS on a Mac (OpenBUGS does not run under Mac OS). The students should bring their own (or their lab窶冱) laptop to be able to carry out possible analyses using R.

June 27. 16:50-18:20, Room 202

Yamagishi hosts a Q & A session for the June 20 activities

Follow-up sessions, schedule TBA

Yamagishi hosts supplementary Q & A sessions. The students may raise questions in Japanese.

No specific requirement is assumed. Familiarity with elementary knowledge of statistics, cognitive psychology and/or behavioral economics will help.

Each student should bring the laptop to the workshop room. The Window 7 computer should have the R, OpenBUGS and JAGS installed. Macintosh users should install R and JAGS software.

By June 7, 2012 Miyamoto will upload a webpage for the workshop to https://faculty.washington.edu/jmiyamot/bayes/bstats.htm. This webpage will show how to install the R, OpenBUGS and JAGS software on a Windows 7 computer, or the R and JAGS software on a Mac (OpenBUGS does not run under Mac OS, but JAGS is functionally very similar to OpenBUGS; it is even possibly better than OpenBUGS). It will also point the reader to background for the workshop.

None. Keep in mind that all instructions will be given in English, and the term paper is to be written in English as well.

A short written assignment shall be required.

yamagishiﾂｼhum.titech.ac.jp