Meta Data Scientist interview questions

The interview process is overall extremily smooth -- the recruiters in particular are very helpful and give you lots of information about what to expect. The interview overall is very focused on business acumen as opposed to technical skills. This reflects the role at Meta where your most important asset is using data to influence business decisions -- which often doesn't require anything beyond asking good questions and producing some good summary statistics. There were 4 half hour parts: 1. A "very broad" business problem. Such as "how do you evaluate the success of Newsfeed?" 2. A somewhat narrower but still open ended problem such as "what do you do about 'fake' friend requests"? 3. A SQL section -- "how would you query X from this table?" 4. A stats teaser -- "if a doctor has a test with 99% accuracy but only 5% of the population has a disease, what is the chance their patient has the disease with a positive test?" (Bayes Rule)