WorldQuant Intern Quantitative Researcher Interview Experience — Round One, All Behavioral and Math, No Coding

Hiring process
There are three rounds of interviews in total. This is round one.

Behavioral questions

• Self-introduction
• Why WorldQuant?
• Why do you want to do Quant?
• Introduce your most relevant project
• Future plans
• Vague goals vs. specific goals

Math questions

First question: the child-birth problem
A country has a rule: every family keeps having children until the last child is a boy, then they stop.
Question: what's the final ratio of boys to girls?

Answer
As the number of families approaches infinity, the ratio of boys to girls approaches 1:1.
This can be shown by:

• Every family ends up with exactly one boy
• The expected number of children per family is 2
• So on average there's one girl left over
• So in the end the total number of boys equals the total number of girls

Second question: three boxes
Three boxes contain:

• red, red
• red, blue
• blue, blue You randomly pick a box. The first draw is a red ball, not replaced. The second draw is also a red ball. Question: what's the probability both red balls came from the red-red box? (Classic Bayes probability question.)

Third question: record maximum
Randomly permute 1 to 100.
Iterate through the array.
Maintain a current maximum.
Every time a new maximum appears, you replace it once.
Question: what's the expected number of replacements?

Follow-up
What if the array is 1 to 98, then 100, 100 — what's the expected number then?

Overall impression
The overall interview experience was good, both VPs were pretty friendly.
Behavioral questions and math questions each took up roughly half the interview. No coding was tested, and they didn't ask much about machine learning.
The math questions mainly covered:

• Probability
• Expectation
• Bayes
• Random permutations (record maximum)
• Logical reasoning Medium difficulty — if you've already ground through the classic green-book problems, you've probably seen similar ones before.