Pinterest Data Scientist Interview Experience — A Restaurant Reservation Coding Question and a Deep-Dive Stats Round

I felt like the whole interview experience was really good — the interviewers listened carefully, and there wasn't that "gotta beat you down" attitude some interviewers have. Two of the rounds felt somewhat difficult to me.

coding: Not counting the resume intro, small talk, and the interviewer showing up late, the actual working time was about 30 minutes. The question involved a nested dictionary — I'd seen something similar on the forum before, an airline reservation question. This one was similar. The question was about figuring out what times a restaurant has open for reservations, given the restaurant's business hours and its existing reservations.

stats: The interviewer started with basic distributions. I expected a lot of A/B testing stats questions, but it turned out to be nothing like that. There was one thread that built out from a single question: say you ran a survey, for example collecting satisfaction ratings on some feature, and the male/female ratio among respondents ended up imbalanced. The questions were: 1) is this a concern or not, 2) how would you validate it, 3) what would the female/male distribution look like, and 3) how do you find a probability, e.g., getting 30 females out of 70 respondents. Later they also brought in stratified sampling.