##### Scenario You are a data scientist advising a product team on statistical analysis and experimental design. ##### Question Explain Simpson’s paradox and illustrate it with a concrete numeric example understandable to a non-technical PM. In a forest with three bird species, propose and define metrics that quantify how segregated the species are from one another. A simple linear regression is run to estimate the effect of YouTube ad impressions on product sales. What potential statistical problems do you see and how would you redesign the study? Name ways to reduce a survey’s margin of error. If sample size and confidence level cannot change, what else could you do and why might bootstrap methods help? ##### Hints Discuss confounding, omitted‐variable bias, experimental control, effect size, confidence intervals, and resampling logic in plain language.

This interview question evaluates statistical assumptions, formulas, estimation strategy, uncertainty, edge cases, and interpretation in a realistic interview setting. A strong answer for Understand Simpson's Paradox with Simple Examples states assumptions, handles edge cases, explains trade-offs, and shows how to validate the result clearly.

How do I approach Statistics & Math interview questions?

Statistics & Math questions require understanding of core concepts and practice. PracHub provides solutions with explanations to help you master statistics & math interviews.

What difficulty level is this interview question?

This is a medium difficulty Statistics & Math question, commonly asked during Onsite rounds at Google.

What role is this question designed for?

This question is commonly asked for Data Scientist candidates at Google during technical interviews.

Understand Simpson's Paradox with Simple Examples

Scenario

You are a data scientist advising a product team on statistical analysis and experimental design.

Tasks

Simpson’s paradox

Explain Simpson’s paradox in plain language and give a small numeric example that a non-technical PM can follow.

Species segregation in a forest

In a forest with three bird species, propose and define clear, quantitative metrics that capture how segregated the species are from one another. State the minimal assumptions (e.g., grid cells with presence/absence or point locations of birds) and how to interpret each metric.

Ads and sales: regression vs. causality

A simple linear regression is run to estimate the effect of YouTube ad impressions on product sales. What statistical problems might arise? How would you redesign the study to estimate causal lift? Be explicit about confounding, omitted‐variable bias, experimental control, effect size, and confidence intervals.

Survey margin of error

Name ways to reduce a survey’s margin of error (MoE). If sample size and confidence level cannot change, what else could you do, and why might bootstrap methods help?

Hints

Touch on confounders, omitted‐variable bias, experimental control, effect sizes, confidence intervals, and resampling (bootstrap) logic in plain language.

Constraints & Assumptions

Preserve the scope, facts, inputs, and requested outputs from the prompt above.
If the prompt leaves a detail unspecified, state a reasonable assumption before relying on it.
Keep the answer interview-ready: concise enough to present, but concrete enough to implement or evaluate.

Clarifying Questions to Ask Guidance

Clarify the random variables, distributional assumptions, independence assumptions, and desired output.
Show enough derivation for the interviewer to follow the reasoning.
Explain how you would validate the result with simulation or sensitivity checks.

What a Strong Answer Covers Guidance

A correct setup with definitions, formulas, and boundary conditions.
A step-by-step derivation or estimation plan.
Interpretation of the result, including uncertainty and practical limitations.
Checks for assumptions, edge cases, and numerical stability.

Follow-up Questions Guidance

How would the result change if the assumptions were relaxed?
Can you verify the answer with a simulation?
What is the most likely source of estimation error?