Behavioral Case: Executing a 4–6 Hour Take‑Home Data Science Assignment

Context

You are a candidate for a Data Scientist role. You receive a one‑week take‑home with no single correct answer and a suggested 4–6 hours of effort. Deliverables may be slides or a written document plus code. Stakeholders may be unfamiliar with modeling.

Prompts

1. Day 1: How do you scope the problem, time‑box analysis, and proactively align expectations with the recruiter/hiring manager?
2. Deliverable choice: How do you choose between slides vs. a written doc for stakeholders unfamiliar with modeling, and what specific narrative/visuals do you include to make trade‑offs and limitations clear?
3. Execution within the timebox: How do you balance data cleaning, modeling, and visualization to tell an end‑to‑end story? What do you explicitly de‑scope first if time runs short and why?
4. Process timing: If you receive another offer mid‑process, how do you communicate professionally to request an expedited decision without pressuring the team?
5. Code/package hygiene: How do you make your code reviewable and reproducible (structure, environment, seeds, data contracts) and handle follow‑up questions after submission?