The talk is 4-5pm in Room 200 on the 20th floor of the Presbyterian Hospital Building, Columbia University Medical Center.

I’m not sure what I’m gonna talk about. It’ll depend on what people are interested in discussing. Here are some possible topics:

– The failure of null hypothesis significance testing when studying incremental changes, and what to do about it;

– Cool stuff in Stan;

– Bayes for big data: Challenges and possible research directions;

– Multilevel models and N=1 studies: Bridging between big and little data;

– Bayesian workflow;

– Improving statistical graphics through statistical modeling, and vice versa;

– Informative prior distributions and how to be a frequentist;

– Teaching statistics.

Or whatever else comes up.

Cool! In my experience working with medical/pharma people for about a year now, Bayesian workflow is something that is very foreign to a lot of them and may therefore be the most illuminating.

Medical/pharma people might need this more “informative prior distributions and how to be a frequentist”.

What do you mean by “how to be a frequentist”?

Bryan:

Here’s an example of how to be a frequentist.

Informative prior distributions and how to be a frequentist? That’s interesting. What’s that about?

I know you are a professional, but I look at the (partial) list of subjects you can speak on for an hour with a day’s notice, and weep.

Torquemada:

Over the years I’ve accumulated hundreds of examples and hundreds of different things to say, so at this point all I really have to do is start talking, watch the expressions on people’s faces as I speak, and say what comes next. I just have to remember to emote and show some enthusiasm when I talk.

“…all I really have to do…” and “I just…” are not helping.