Ep21 "Disentangling Causation and Correlation" with Guido Imbens
All Else Equal: Making Better Decisions
00:00
Is Running an Experiment a Good Business Decision?
In general I think you want to see running an experiment as just another business decision. An experiment is really just an investment where you buy some information in order to improve your decision making. And if there's a lot of decisions for which that type of information is very helpful then it may be worth setting up a platform where you can do experiments more regularly and do them more effectively. So Gito talking about that random experiments can be very costly we also spoke about the idea of using naturally occurring randomness to try to tease out causality. Do you have any examples in your own experience of where a naturally occurringrandomness has been used by a business to infer causality?
Transcript
Play full episode
