Bayes Theorem Quant Interview Questions
How to solve Bayes theorem questions in quant interviews with base rates, likelihoods, examples, and common traps.
Candidates who miss conditional probability and base-rate questions.
What Bayes questions test
Bayes questions test base-rate discipline. Interviewers want to see whether you update a prior with evidence instead of reacting only to the headline likelihood. The safest setup is a two-by-two table or a tree of true state and observed signal.
Concrete example
Suppose 1 percent of candidates have a trait and a test is 90 percent accurate both ways. A positive result is not 90 percent likely to be true. Out of 10,000 people, about 90 true positives and 990 false positives appear, so the posterior is about 8.3 percent.
How to practice
Convert percentages into counts. Counts reduce algebra errors and make the answer easier to explain. Then practice variants where the base rate changes, the signal has asymmetric accuracy, or there are multiple sequential signals.
Common mistakes
The common trap is ignoring false positives when the base rate is small. Another is using P(signal | state) as if it were P(state | signal). Say both events out loud; if the sentence changes, the probability changed too.
Practice the pattern
Use the LeetQuidity curriculum and calibration to turn this topic into a focused practice plan.