Coding part: it was bootstrap + permutation coding. They give you a dataset to code on, it has to be bug free, and there are some follow-up coding questions. I wrote 3 problems and I feel like there should have been more. (They make you actually run it, it's not just writing out some simple logic.)
Case part: the interviewer shared a PPT and went through it with me. It was mainly about sizing, how to do the sizing, which of the various visualization charts is good and which is bad, and how to set up an experiment.
Stats part: they asked about Diff in Diff, with a specific case. (You have to study DiD!! I can't stress it enough, I'm saying it several times.)
HM round: that round was really weird, it actually had a case in it too (a question about engagement), not BQ (and very team specific). I was surprised.
Good luck everyone!!!
Discussion
Loading comments…