Calculate Expected Day for First Selection in Sampling

Quick Overview

Meta probability prompt comparing expected first-selection day under sampling without replacement versus with replacement, using random permutations, uniform day distribution, and geometric waiting time.

Calculate Expected Day for First Selection in Sampling

Company: Meta

Role: Data Scientist

Category: Statistics & Math

Difficulty: easy

Interview Round: Onsite

##### Scenario Daily sampling of 10 names from 1 000 people ##### Question Without replacement (no repeats until everyone is chosen), what is the expected day on which a given person is first selected? With replacement (names can repeat each day), what is the expected day on which that person is first selected? ##### Hints The no-replacement waiting time is uniformly 1…100; with replacement it follows a geometric distribution with p = 0.01.

Quick Answer: Meta probability prompt comparing expected first-selection day under sampling without replacement versus with replacement, using random permutations, uniform day distribution, and geometric waiting time.

|Home/Statistics & Math/Meta
Meta logo
Meta
Jul 12, 2025, 6:59 PM
easyData ScientistOnsiteStatistics & Math
27
0

Expected Day of First Selection in Daily Sampling

There are 1,000 people. Each day, 10 distinct names are selected uniformly at random. Day counting starts at 1.

Compare two regimes:

  1. Without replacement across the whole campaign: once a person is selected, they are not eligible again until all 1,000 have been selected. This is equivalent to a random permutation split into 100 days of 10 people each.
  2. With replacement across days: each day is a fresh draw of 10 distinct names from the full 1,000, so the same person can be selected on multiple days.

Constraints & Assumptions

  • The person of interest is fixed before the sampling starts.
  • In the with-replacement case, daily selection events for that person are independent across days.
  • Report expected day, not probability of selection by a given day.

Clarifying Questions to Ask Guidance

  • Does "without replacement" apply only within a day, or across the whole campaign?
  • Are we choosing exactly 10 distinct people each day?
  • Does day 1 count as the first possible selection day?

What a Strong Answer Covers Guidance

  • Without replacement, the person's position in the random permutation is uniform from 1 to 1,000.
  • Grouping positions into 100 days of 10 makes the first-selection day uniform on days 1 through 100.
  • The expected day without replacement is (1 + 100) / 2 = 50.5 .
  • With replacement across days, daily selection probability is 10 / 1000 = 0.01 .
  • The first-selection day follows a geometric distribution with mean 1 / 0.01 = 100 .
  • Clear explanation of why the two regimes differ.

Follow-up Questions Guidance

  • What is the probability of being selected by day 30 in each regime?
  • How would the expected day change if 20 names were drawn each day?
  • What if a person could be selected more than once within the same day?
Loading comments...