Ben Kuhn

Berkeley, California

I'm currently a data scientist at an early-stage financial technology startup.

Previously, I helped start Harvard College Effective Altruism, no...

In my opinion, many people in the EA community make sub-optimal decisions because they either don't realize they could apply various statistical tools, or don't know them well enough to use them properly.

I can help with this. If you're doing an EA project and think that you have a problem that would be easier if you understood statistics better, email me and I'll give you feedback! And more in-depth help as my time permits.

For instance, here are some example problems I could help with:

  • How trustworthy is this paper's findings?
  • What does the academic literature say about this phenomenon?
  • I have some data. Can I do anything interesting with it?
  • How can I predict some outcome, rigorously and without bias?
  • How can I improve this experimental design?
  • Should I run an experiment at all, if I can afford a sample size of N?
  • I just received some information. How strongly should I update on it?