StanCon is one of the least intimidating places to present on applied statistics. I mean, @Bob_Carpenter has an entire case study on modeling baseball statistics. So you really can present just about anything with Stan in it.
We try very hard to keep the tone of these forums, conferences, etc. constructive.
I agree StanCon is a great forum (obviously :-) ), but to me sports are almost the prototypical overused domain in statistics. So to underscore that you can definitely present anything, we’ve already had great presentations involving
cell metabolism, HIV prevalence, clinical trials, galaxy clusters, neutrino mass, drug dynamics, cocoa beans, gas consumption, fairness, police statistics, communicating statistical results and much more.
And we would love to see more topics, so if your work involves statistics in the context of knitting, art, video games, psychological models, ecology, construction, classical music, plush toys, handshaking, the Internet, philosophy, ethics, transcendental experiences, community building, governance, child play, dreams, dragons, unicorns or literally anything else, consider presenting your work!
To me the best part of StanCon is learning how people in hugely different background from mine approach problems that have some connection to my work.
StanCon 2020 registration is now open!
Please find details on tutorials proposals, thematic session proposals, and talk proposals at https://mc-stan.org/events/stancon2020/#tutorials.
Would you all consider having a session on “open problems” where folks from different disciplines can present mini-presentations or posters on problems that they think would benefit from a Bayesian approach (and try to inspire interest)? @charlesm93 and I were talking about how this might be a nice thing to add…
I like the idea. Potentially can be a piggyback of the poster session. Could you email firstname.lastname@example.org for a better coordinated discussion? Not every organizer is active on discourse.
Sure thing! I’ll loop Charles in too.
This is similar to a proposal I sent to @syclik a while ago. My focus was slightly different, but I think your take has also a lot of value. There probably is a best of both worlds approach somewhere. Here’s my version, will be happy to help get this off the ground in any form:
Problem solving session: people can submit a new type of submission a “Problem statement”, describing a problem they are (unsuccessfully) trying to solve and why it is important. During the actual session, participants form small groups around the individual problems (15 people max), the submitter presents a short description of the problem and the group then spends some time in a structured (facilitated) discussion, trying to chart possible lines of attacking the problem.
This is a great idea for both posters and as a Thematic session proposal. If you gather a few folks/topics we can start adding information in case other people are interested having a joint session.
I agree that sports examples can be overused–indeed, I’m on record saying that teachers should use examples from a wide range of application areas. But there are good reasons to use examples from sports, as one of many domains of application.
Sports examples can be good in the same way that psychometrics examples can be good: they combine various aspects of business and human performance studies, in relatively non-technical contexts that can many students can relate to.
In general, I think education examples are better than sports examples, because not everyone is interested in sports, but just about all students are familiar with education, testing, etc.
But I think sports are a better example than, say, gambling, in that the modeling in sports can be relevant for a larger class of other problems. When I was a student, many years ago, probability textbooks were full of examples of poker hands, dice games, etc. This always seemed like a wasted opportunity to me: students end up learning irrelevant things like the rules of the game of craps . . . seems like a dead end to me. In contrast, our World Cup example demonstrates general principles of models in psychometrics, and our golf example demonstrates a very simple example of modeling using geometry.
I also like examples in particular subjects such as economics, political science, or genetics. Sometimes these examples require learning some subject-matter details (for example, rules for gene combinations), but at least then the student is learning some real science along the way.
Finally, I agree with your general point that it’s good to see a wide range of examples. I don’t think sports or political science are overused domains–of course I’d say that!–but in any case there’s room for more topics.
Can I still apply for a scholarship?
Yes, you can!
The StanCon organizing committee just put up an announcement here:
Hi all. At the Stan meeting today we discussed the possibility of an online StanCon this summer, as this would be an opportunity to engage thousands of users. We were wondering how to coordinate this with the previously planned Oregon State University StanCon: maybe the Oregon meeting would be postponed for a year?
Will there be an opportunity to register for the online version? Or will it be for those who registered for the OSU edition?
Any update on Stancon plans (in person or remote) this summer?
The committee will meet on the coming Tue to discuss our options. We’ll make announcement as soon as decision is made.
Any update? Thanks again.
Just posted the update here: https://discourse.mc-stan.org/t/stancon-2020-going-virtual
TL;DR: StanCon 2020 is going virtual!