← Meta Interview Insights

Meta·Data Scientist·Technical Phone Screen·Intermediate

Intermediate
Apr 2026

Summary

Probability and Bayes question for a Meta data scientist role. Two parts, straightforward setup but the conditional probability follow-up is where things get interesting.

Questions Asked (2)

Q1

Two rooms have the following occupancy priors: probability 1/3 both are occupied, probability 1/3 exactly one is occupied, probability 1/3 both are empty. You pick a room uniformly at random to check. What is the probability the room you check is occupied?

Product Analytics & Metrics
Author's notes

Part (a) is just total probability.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Use the law of total probability by conditioning on the occupancy state (both occupied, exactly one occupied, both empty). For each state, compute the probability that a randomly chosen room is occupied, then take the weighted average using the given prior probabilities.

Pro tip: After computing the answer, briefly mention that this is a simple application of conditional expectation and that the result aligns with intuition: on average, half the rooms are occupied, so the probability is 1/2.

1. Define the events

Let O be the event that the checked room is occupied. Define the occupancy states: A = both occupied, B = exactly one occupied, C = both empty.

2. Assign prior probabilities

Given P(A) = 1/3, P(B) = 1/3, P(C) = 1/3.

3. Compute conditional probabilities

For each state, find P(O|state): P(O|A) = 1 (both occupied), P(O|B) = 1/2 (one occupied out of two), P(O|C) = 0 (both empty).

4. Apply law of total probability

Compute P(O) = P(O|A)P(A) + P(O|B)P(B) + P(O|C)P(C) = 1*(1/3) + (1/2)*(1/3) + 0*(1/3) = 1/3 + 1/6 = 1/2.

5. Interpret the result

The probability is 1/2, meaning there is a 50% chance the checked room is occupied.

Key Points to Mention

  • Law of total probability
  • Conditioning on occupancy states
  • Uniform random selection of room
  • Weighted average of conditional probabilities
  • Intuition: expected occupancy is 1 room out of 2, so probability is 1/2
  • Independence of room selection from occupancy state

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q2

Given that the room you picked is occupied, what is the probability the other room is also occupied? Walk through the Bayes calculation explicitly.

Product Analytics & MetricsRoot Cause Analysis
Author's notes

This is where it gets trickier.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

First, clarify the problem setup and define the events and prior probabilities. Then apply Bayes' theorem step-by-step, explicitly stating the formula and computing each term. Finally, interpret the result in the context of the problem and discuss any assumptions or implications.

Pro tip: Always state your assumptions clearly, especially about independence and prior probabilities, as this shows rigor and prevents misunderstandings. Also, relate the calculation to a real-world scenario to demonstrate practical insight.

1. Clarify the problem and assumptions

Restate the question in your own words and specify the events: e.g., let A be the event that the other room is occupied, and B be the event that the chosen room is occupied. State any assumptions about independence or prior probabilities.

2. Define prior probabilities

Assign prior probabilities P(A) and P(B) based on the given context or reasonable assumptions. If not given, discuss how you would estimate them (e.g., from historical data or symmetry).

3. Apply Bayes' theorem

Write Bayes' formula: P(A|B) = P(B|A) * P(A) / P(B). Compute the numerator and denominator, using the law of total probability for P(B) if needed.

4. Compute and interpret the result

Plug in the numbers to get the posterior probability. Interpret the result in plain language, explaining what it means for the likelihood that the other room is occupied given that the chosen room is occupied.

5. Discuss extensions or sensitivity

Mention how the result changes with different priors or assumptions, and relate it to similar problems in product analytics or root cause analysis.

Key Points to Mention

  • Bayes' theorem formula and its components
  • Definition of events and conditional probability
  • Assumption of independence (or lack thereof) between rooms
  • Prior probabilities and how they are determined
  • Law of total probability for computing the denominator
  • Interpretation of the posterior probability in context

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.