It’s for hierarchical modeling for policy analysis in Stan.

We’re really excited about this project. Will share more details soon, but wanted to get this out right away.

Posted by Andrew on 23 October 2019, 9:50 pm

It’s for hierarchical modeling for policy analysis in Stan.

We’re really excited about this project. Will share more details soon, but wanted to get this out right away.

Filed under Bayesian Statistics, Causal Inference, Decision Theory, Economics, Multilevel Modeling, Stan

Comment (RSS)
| Permalink
## Recent Comments

- jim on (1) The misplaced burden of proof, and (2) selection bias: Two reasons for the persistence of hype in tech and science reporting
- Mikhail Shubin on “Pictures represent facts, stories represent acts, and models represent concepts.”
- Mikhail Shubin on “Pictures represent facts, stories represent acts, and models represent concepts.”
- Stephen Olivier on “this large reduction in response rats”
- Phil on Bill James is back
- Michael Nelson on “Pictures represent facts, stories represent acts, and models represent concepts.”
- Daniel Lakeland on (1) The misplaced burden of proof, and (2) selection bias: Two reasons for the persistence of hype in tech and science reporting
- Ron Kenett on “Pictures represent facts, stories represent acts, and models represent concepts.”
- Andrew on Birthday analysis—Friday the 13th update, and some model checking
- Talia Page on Birthday analysis—Friday the 13th update, and some model checking
- jim on (1) The misplaced burden of proof, and (2) selection bias: Two reasons for the persistence of hype in tech and science reporting
- jim on (1) The misplaced burden of proof, and (2) selection bias: Two reasons for the persistence of hype in tech and science reporting
- jonathan on “Pictures represent facts, stories represent acts, and models represent concepts.”
- Rudy Banerjee on “Pictures represent facts, stories represent acts, and models represent concepts.”
- Daniel Lakeland on (1) The misplaced burden of proof, and (2) selection bias: Two reasons for the persistence of hype in tech and science reporting
- Carlos Ungil on (1) The misplaced burden of proof, and (2) selection bias: Two reasons for the persistence of hype in tech and science reporting
- H Smith on Postdoc in Bayesian spatiotemporal modeling at Imperial College London!
- Daniel Lakeland on (1) The misplaced burden of proof, and (2) selection bias: Two reasons for the persistence of hype in tech and science reporting

## Categories

The start date’s some time early 2020.

We’ll also be fitting models involving equilibrium conditions specified with differential and algebraic equations. Ideally, we’d like to extend to stochastic versions of these.

We’re hoping to find someone who wants to work with us and a great group of economists fitting Bayesian models and then describing them in papers and reproducible case studies. Developing reusable packages for economics applications would also be great as would contributing to core Stan packages. There is also the latitude to engage in your own research within the broad framework of the Stan project and Bayesian econometrics.

This work will be done in conjunction with a computational post-doc or research scientist working on related Stan infrastructure.