Dear all,

A relevant postdoc for political researchers. Grateful if you can spread
the word!

The postdoc in this project will work with the PI and manage at least one
research assistant in developing a system for comparing underlying
theoretical and causal assumptions from statistical models in the published
literature. This will involve starting with a subfield where weak theory is
a pervasive problem: in macro-comparative public opinion and social policy
research. Causal models will be extracted from the existing research, and a
system to compare (meta-analyze) causal models and to identify where causal
assumptions are not compatible will be developed.

The applicant should be able to demonstrate machine learning knowledge
using Python and ideally experience working with R, able to work fluently
in the project language of English.

Other assets would be knowledge of causal inference, reproducible
workflows, virtual machine environments, mathematical and/or philosophical
logic, git, dokku, Google ColabPro and familiarity with general
sociological or political science research on opinion, attitudes and social
policy. The ideal candidate would be ready to learn diverse new skills and
implement machine algorithms for a variety of purposes such as
classification, prediction and extraction of information. This project aims
to promote open science with ethical and reproducible research.

https://www.uni-bremen.de/en/university/the-university-as-an-employer/job-vacancies-1/job/1432

Best,

Nate Breznau
*sent remotely

**************************************************************
               Political Methodology E-Mail List
   Editors:  Dominique Lockett and Luwei Ying
                  <[log in to unmask]>
**************************************************************
     Send messages to [log in to unmask]
  To join the list, cancel your subscription, or modify
                 your subscription settings visit:

  https://www.cambridge.org/core/membership/spm/mailing-list

**************************************************************