Something on Applied Bayesian Statistics
April 27, 4:10-5 p.m., 1011 Evans Hall
I will deliver one of the following three talks:
1. Of beauty, sex, and power: Statistical challenges in estimating small effects
2. Why we (usually) don’t worry about multiple comparisons
3. Parameterization and Bayesian modeling
Whoever shows up on time to the seminar gets to vote, and I’ll give the talk that gets the most votes.
Sex sells: let's go for 55%, 20%, 25%
This is just a secret ploy to try and get academics to show up for something on time….
Let me know how it works. Could prove to be an excellent strategy.
Which voting system will you use?
Needed to understand slides for #3 "Parameterization and Bayesian modeling" — our host's "Handy statistical lexicon"
http://www.stat.columbia.edu/~cook/movabletype/ar…
The Folk Theorem: When you have computational problems, often there's a problem with your model.
The Pinocchio Principle: A model that is created solely for computational reasons can take on a life of its own.
You should have someone spin a wheel to decide, Wheel of Fortune style. The band Yo La Tengo are currently doing this for the first set of their tour.
http://www.matadorrecords.com/matablog/2011/03/30…
Oh man, I wish I would've learned about this sooner! I'm going to try really hard to make it over, and if I do I'm going to vote for the multiple comparisons talk.
You need to come to Southern California to do a talk one day.
Please.
Dear Professor Gelman,
I went to your talk today! I was so impressed! I am happy that you came all the way to Berkeley to give a presentation. Thank you for talking with me today. I am very grateful that we were able to share our memories of Seth.