Data Science Intern Interview Questions: Stats, SQL, ML and Case Rounds

Data science internships test three things in a short loop: whether you can reason about data statistically, whether you can write the SQL and Python to get at it, and whether you can connect an analysis to a decision. This page covers the rounds to expect, the questions that come up with what a strong answer contains, two sample answers, what interviewers expect from an intern, the mistakes that cost offers, and a two-week prep plan.

What the data science intern interview looks like

Questions that come up, and what strong answers contain

1. Explain a p-value to a non-technical manager. A plain-language definition, what it does not mean, and how you would use it in a decision.
2. How would you design an A/B test for a new feature? Hypothesis, metric, randomization unit, sample size and power, run length, and what would make you stop early.
3. Write a query for seven-day retention by signup week. Correct joins and date logic, and a word on how you would check the result.
4. What is the difference between precision and recall, and when does each matter? Definitions plus an example where a false positive costs more than a false negative, or the reverse.
5. How do you handle missing data? Why it is missing first, then options such as dropping, imputing or flagging, and their risks.
6. Walk me through a project you are proud of. The question, the data, your method, the result and what you would change.
7. A metric dropped 10 percent overnight. What do you check? Data issues first, then segments, then external causes, in a clear order.

Two sample answers

Walk me through a project you are proud of.

For a class project I predicted which campus dining items would sell out. I cleaned two semesters of sales data, built a gradient-boosted model and compared it with a simple day-of-week baseline. The model cut prediction error by about a third, and the dining hall used my weekly forecast to adjust orders. If I did it again I would add weather data earlier, because it explained most of the misses.

How would you design an A/B test for a new feature?

I'd start with the decision we want to make and one primary metric, say seven-day retention. I'd randomize by user, estimate the sample size from the smallest lift worth detecting, and fix the run length up front to avoid peeking. I'd also track a guardrail metric like app crashes, and segment results only after the main read.

What interviewers expect from an intern

• Sound statistics fundamentals over advanced models.
• SQL you can write without looking things up.
• The ability to explain results to someone who does not do data science.

For the full-time version of these questions, see data scientist interview questions.

Mistakes that cost offers

• Jumping to a model before understanding the question.
• Reciting definitions without an example.
• Writing SQL that ignores duplicates or NULLs.
• Describing group projects without saying what you personally did.

A two-week prep plan

1. Days 1 to 4: SQL every day, including joins, window functions and date logic.
2. Days 5 to 7: probability and statistics, especially hypothesis tests and A/B testing.
3. Days 8 to 10: pandas practice and one metrics case per day out loud.
4. Days 11 to 14: rehearse two project walkthroughs and run two mock interviews.

Timing matters as much as preparation: see our internship recruiting timeline, and once you are in, turning an internship into a return offer. Engineering candidates should also read software engineer intern interview questions.

Where CoPilot Interview fits

Intern interviews reward clear explanations of your own projects and coursework, and that takes practice out loud. CoPilot Interview is an AI interview assistant; its free AI mock interview builds data science intern questions from your résumé and the job description, reads them aloud, scores each spoken answer out of ten and shows a model answer to compare. The free plan gives 15 minutes a day of live help in sessions of up to 15 minutes, with AI mock interviews included, with no card, which suits a student budget. For real interviews and online assessments, follow your employer's rules.

Disclosure: CoPilot Interview is our product, so weigh what we say about it accordingly.

Frequently asked questions

What SQL is asked in data science intern interviews?

Joins, aggregation with GROUP BY, window functions for rankings and running totals, and date logic for retention or cohorts.

Do data science intern interviews include machine learning?

Usually basics: bias and variance, overfitting, common metrics and when you would use a simple model. Deep theory is rarer at intern level.

How should I talk about class projects?

Like real work: the question, the data, your method, the result and what you would improve.

Can I use AI tools during an internship interview?

Only where the instructions allow it. Online assessments and many live rounds forbid outside help. Use AI to practice beforehand, and follow your employer's rules.