Compute averages and binomial/Poisson probabilities

Quick Overview

This question evaluates proficiency in streaming summary-statistics updates and probability calculations, specifically computing an updated mean from aggregated summaries and comparing Binomial versus Poisson probability computations within the Statistics & Math domain.

Compute averages and binomial/Poisson probabilities

Company: Boston Consulting Group

Role: Data Scientist

Category: Statistics & Math

Difficulty: medium

Interview Round: Take-home Project

Part A (streaming mean): A dataset has 1,000 observations with mean 12.4. You remove 50 observations whose mean is 20.1 and then append 30 new observations whose mean is 9.8. Without recomputing from raw data, what is the exact new mean? Show the formula you use and the final value to 3 decimal places. Part B (binomial vs Poisson): Let X ~ Binomial(n=40, p=0.07). 1) Compute P(X=0), P(X≤2), and P(X≥3) exactly (4 decimal places). 2) Approximate the same three quantities using a Poisson(λ = n·p) model with continuity where appropriate; report absolute errors for each quantity (|exact − approx|). 3) Briefly justify when the Poisson approximation is appropriate here and how you would tighten the approximation if p were larger.

Quick Answer: This question evaluates proficiency in streaming summary-statistics updates and probability calculations, specifically computing an updated mean from aggregated summaries and comparing Binomial versus Poisson probability computations within the Statistics & Math domain.

|Home/Statistics & Math/Boston Consulting Group
Boston Consulting Group logo
Boston Consulting Group
Oct 13, 2025, 9:49 PM
mediumData ScientistTake-home ProjectStatistics & Math
8
0

Streaming Mean and Binomial vs Poisson Approximation

Part A — Streaming Mean Update

You have an existing dataset of N = 1,000 observations with mean 12.4. You:

  • Remove r = 50 observations whose mean is 20.1, then
  • Append a = 30 new observations whose mean is 9.8.

Without recomputing from raw data, compute the exact new mean. Show the formula you use and report the final value to 3 decimal places.

Part B — Binomial vs Poisson Approximation

Let X ~ Binomial(n = 40, p = 0.07).

  1. Compute P(X = 0), P(X ≤ 2), and P(X ≥ 3) exactly. Report each to 4 decimal places.
  2. Approximate the same three quantities using a Poisson(λ = n·p) model (use continuity only if appropriate), and report absolute errors for each quantity (|exact − approx|).
  3. Briefly justify when the Poisson approximation is appropriate here, and how you would tighten the approximation if p were larger.
Loading comments...