We will use this space each week in the news blast for Justin to share new ideas he thinks may be of interest for our behavioral & experimental community. If you ever have ideas for topics, please share your ideas with
Justin
.
This Week's BRITE Idea:
Last week we shared an interesting paper that used machine learning techniques to help explore which potential behavioral models (using heuristics in that case) best fit the data and argued that could be an increasingly useful approach even in experimental studies.
Machine learning is not at the core of experimental studies, because we are often interested in estimating a treatment effect and the comparison of simple means does that. But when it comes to using measurements from experiments (e.g., preference measurements) and model building, predictive techniques could be useful.
This week we want to highlight some resources for those who are thinking of moving in this direction. Justin is interested in this and thinking of ways to invest in doing it himself, but is not yet up to speed. Anyone interested in possibly creating a reading/practice group on this over the summer, maybe we should discuss?
Okay, the resources: Here is a link to a
summer workshop series
put on at the NBER in 2015. Here is an
approachable paper
in the Journal of Economic Perspectives on the topic of machine learning for economists. (Side note: one of the authors Jan Spiess will be giving a job talk in the econ department in a few weeks). Finally, I just saw
this page
put together with links on places to get started with ML methods.
