I am a second year doctoral student in the Biostatistics Department at UNC. Recently I have been developing methods for determining dynamic treatment regimes capable of incorporating multiple competing outcomes such as efficacy, utilities (e.g. from a patient or a policy maker), cost, and side effect burden which are interpretable to domain practitioners.
PhD Student,
University of North Carolina
Post-baccalaureate Studies,
University of Pennsylvania
BSc in Policy & Management, 2012
Carnegie Mellon University
Tomorrow I’ll be giving an introductory talk/training session on using Amazon’s Mechanical Turk for conducting survey based research to a group of new research assistants here at the university of Pennsylvania. I’ve posted the talk here (which is still a rough draft as of the time of this posting), and the source is available on Github. I plan on expanding this is into more detailed How-To guides in the future because I think Amazon’s documentation is poor, and while there is good documentation targeted at psychology/social science researchers out there it tends to presume a level of familiarity with Mechanical Turk (e.