Data Scientist Assessment Template

A ready-to-run data science hiring test covering Python, stats, SQL, and ML — with live notebook execution.

Duration
90 minutes
Questions
11
Level
Mid-Level
Passing Score
70%

What this template measures

Every skill needed for a data scientist hire, covered across MCQ, coding, and essay questions.

Statistics

Probability, hypothesis testing, A/B design, confidence intervals.

Python Data Stack

pandas, numpy idioms, plotting with matplotlib.

SQL Fluency

Joins, window functions, analytical queries.

ML Fundamentals

Feature engineering, model evaluation, bias/variance.

Model Evaluation

CV, stratification, metrics appropriate to problem.

Communication

Explains analysis and tradeoffs clearly in writing.

Sample questions from this template

A preview of the questions you'll see when you use this template.

Multiple ChoiceMediumQuestion 1

You run an A/B test and get p=0.04 with n=10,000 per arm. The conversion lift is 0.1%. Which is most accurate?

  • A.Result is statistically significant and practically meaningful
  • B.Result is statistically significant but practically trivial
  • C.Result is not significant at α=0.05
  • D.Cannot determine without raw data
CodingMediumPythonQuestion 2

Given a pandas DataFrame `orders` with columns (user_id, amount, created_at), compute: - Total amount per user - 30-day rolling average of daily order count (across all users) - Top 10 users by total amount

CodingHardPythonQuestion 3

You're building a churn classifier. Given a dataframe with features + binary `churned` target: - Split data with stratification - Handle imbalanced classes - Train a baseline logistic regression - Report accuracy, precision, recall, F1, ROC-AUC on held-out set - Explain which metric matters most for this problem

EssayHardQuestion 4

Design an A/B test to measure the impact of a new pricing page. Cover: hypothesis, success metric, power analysis assumptions, stratification strategy, and stopping rules.

Scoring rubric

How candidates are evaluated on this template.

Dimension
Description
Weight
Statistical Correctness
Uses appropriate tests, interprets results correctly.
30%
Python Fluency
Uses pandas/numpy idioms, avoids anti-patterns.
25%
ML Soundness
Splits, evaluation metrics, and model choice match problem.
20%
SQL Fluency
Queries are correct and efficient.
15%
Communication
Explains tradeoffs and reasoning clearly.
10%

Frequently asked questions

What libraries are preinstalled?+

pandas, numpy, scikit-learn, matplotlib, seaborn, scipy, statsmodels, and jupyter. PyTorch available in ML-focused variants.

Can I customize this template?+

Yes. Every question, time limit, weighting, and rubric dimension is fully editable. Use the template as a starting point and tailor it to your role and seniority level.

Does this template include AI cheat detection?+

Yes. All ClarityHire assessment templates ship with code coherence AI, keystroke biometrics, and paste detection enabled by default. You can dial integrity level per role.

Can candidates see sample questions before starting?+

Yes. Each template supports unscored practice questions so candidates warm up before the real assessment begins. That way you measure skill, not test anxiety.

Launch Your Data Science Assessment Today

Customize this template and invite candidates in minutes.