Data Scientist Mock Interview: 40 Questions to Practice
Data science interviews can combine several disciplines.
You may encounter
statistics machine learning SQL experimentation business cases modeling communication A realistic mock interview should reflect that variety. Statistics
1. What is a confidence interval?
2. What is statistical significance?
3. What is a p-value?
4. What is selection bias?
5. What is variance?
6. What is overfitting?
7. How do you handle missing data?
8. How would you design an experiment?
Machine learning
9. What is the difference between supervised and unsupervised learning?
10. What causes overfitting?
11. How would you choose a model?
12. How would you deal with imbalanced classes?
13. How do you evaluate classification?
14. How do you evaluate regression?
15. What is feature leakage?
16. How would you debug a model that suddenly performs worse?
Modeling decisions
17. When is a simple model better than a complex model?
18. How do you select features?
19. How do you think about interpretability?
20. How do you handle noisy data?
21. How do you validate a model?
22. How would you monitor a model in production?
Business cases
23. How would you predict customer churn?
24. How would you measure recommendation quality?
25. How would you determine whether a product change worked?
26. How would you investigate declining engagement?
27. Which metric would you optimize?
Communication
- Explain a machine-learning concept to a nontechnical stakeholder.
- How do you communicate model uncertainty?
- What would you do if a stakeholder disagreed with your analysis?
- Tell me about your most important analytical project.
- Tell me about a model that failed.
- Tell me about a time you changed your approach.
- Tell me about a difficult stakeholder.
- Tell me about a time data changed your opinion.
- Tell me about a technical disagreement.
- Tell me about an incorrect analysis.
- What area of data science are you currently developing?
- Why do you want this role?
- What kind of data problem interests you most?
Behavioral
Practice the entire reasoning chain
Don't practice
“What is gradient boosting?” only as a definition exercise.
Practice
“You have a churn problem. What would you do?”
Then reason through
objective → data → target → features → model → evaluation → deployment → monitoring This creates interview-ready thinking. The business-impact test
For every project on your resume, be ready to explain
Why did the problem matter? Why did you choose the approach? What changed? How did you measure success? What would you do differently? Those questions often reveal more than memorizing definitions. [Run a Data Scientist mock interview with Mockly]