Interview Questions

Data Scientist Interview Questions

10 questions you are likely to meet in a Data Scientist interview. For each one: what a strong answer covers, and the free lesson that prepares it.

  • 10 Questions
  • 8 Skills
  • Reviewed 16 November 2026

How the Interview Usually Runs

Usually a SQL and Python round, a statistics and machine learning round, a case or take-home analysis and a behavioral round.

Foundations

The concepts interviewers check first.

  1. Explain the difference between correlation and causation with an example.

    What a strong answer covers

    Two variables can move together because of a third factor or by chance. Ice cream sales and drownings both rise in summer. Causation needs an experiment or careful causal methods that control for confounders.

    Prepared InS5 · #58Probability and Statistics, coming in Season 5

  2. How do you design and read an A/B test?

    What a strong answer covers

    Define one primary metric and the minimum effect worth detecting, compute the sample size, randomize users, run for full cycles, avoid peeking, then read the effect with its confidence interval and check guardrail metrics.

    Prepared InS13 · #156Product Analytics, coming in Season 13

  3. Which metric would you choose for an imbalanced classification problem, and why?

    What a strong answer covers

    Accuracy misleads when one class dominates. Use precision, recall, F1 or the precision-recall curve, chosen from the cost of each error, and report results per segment.

    Prepared InS6 · #65Evaluating Models, coming in Season 6

Hands-on

How you would actually build or do it.

  1. How do you handle missing values?

    What a strong answer covers

    First ask why values are missing. Then drop, impute with a simple statistic or a model, or add a missing indicator, fitting any imputation on training data only to avoid leakage, and check the effect on results.

    Prepared InS6 · #66Features and Pipelines, coming in Season 6

  2. Write a SQL query that finds the top 3 products per category by revenue.

    What a strong answer covers

    Aggregate revenue per product with GROUP BY, rank within each category with a window function such as ROW_NUMBER or RANK partitioned by category and ordered by revenue, then filter rank 3 or less in an outer query or CTE.

    Prepared InS4 · #44SQL 3, coming in Season 4

  3. How would you use an LLM to classify 100,000 customer reviews, and check its quality?

    What a strong answer covers

    Write a clear prompt with labelled examples and structured output, label a sample by hand as a golden set, measure agreement, fix the prompt, then run the full set in batches and spot-check results and cost.

    Prepared InS13 · #151Why Evaluation, coming in Season 13

Scenario

Open problems: design, debugging and judgment.

  1. Sales dropped 15% last month. How do you investigate?

    What a strong answer covers

    Check the data first, then break the drop down by region, channel, product and customer segment, line it up with events such as price changes or outages, form hypotheses, test them, and present the most likely cause clearly.

    Prepared InS4 · #50Business Intelligence and Data Storytelling, coming in Season 4

  2. A stakeholder asks you to predict customer churn. How do you approach it end to end?

    What a strong answer covers

    Agree the definition of churn and the decision the model supports, build features from history without leakage, use a time-based split, start with a baseline, evaluate with business costs, and plan how predictions trigger action.

    Prepared InS6 · #66Features and Pipelines, coming in Season 6

  3. How do you explain a single prediction to a business user?

    What a strong answer covers

    Use feature importance and per-prediction contributions such as SHAP values, show the top factors in plain words, compare with similar cases, and be honest about uncertainty and what the model cannot know.

    Prepared InS6 · #67Trees and Boosting, coming in Season 6

Behavioral

Your experience, told with the STAR method.

  1. Tell me about a time your analysis changed a decision.

    What a strong answer covers

    A STAR story: the decision at stake, the data you used, the finding that surprised people, how you presented it, and the measurable result.

    Prepared InS17 · #224Behavioral Interviews and Negotiation, coming in Season 17

Skills Checklist

What Data Scientist interviews test, each with the free lesson that teaches it. Tick what you can already explain.

0 of 8 ready
  • S4 · #44SQL 3

  • S4 · #48pandas

  • S5 · #58Probability and Statistics

  • S4 · #50Business Intelligence and Data Storytelling

  • S6 · #66Features and Pipelines

    What is Machine Learning?
  • S6 · #65Evaluating Models

  • S13 · #156Product Analytics

  • S8 · #87The OpenAI API

    What is a Large Language Model?

Your Next Step

Learn Everything These Questions Test

The Data Scientist path takes you there lesson by lesson, from zero, for free.

Prepare for the Interview, Free

Every ai2hired lesson is built from real AI job ads, and the last season is all about getting hired.