Course contentsShow
← Data Science
Lesson 1205 of 2,14530. Exploratory Data Analysis (EDA) ProcessFree lesson

Pre-Analysis Setup: Documentation and Question Definition

Establish clear objectives, document data sources, and define specific questions before touching the data.

Pre-Analysis Setup: Documentation and Question Definition

You'll learn how to create a disciplined foundation before you touch any data.

The Critical Pre-Work

Before you run a single line of code or open your dataset, you need a documented plan. Pre-analysis setup means writing down what you're trying to learn, where your data comes from, and what questions you must answer—before exploratory work begins.

Think of it like a construction blueprint. You wouldn't start building a house by randomly nailing boards together. Similarly, jumping into data without clarity leads to wandering analyses that find patterns but answer nothing useful.

What to Document Before You Start

1. The Core Question: Write one clear sentence describing what decision or insight this analysis will support. "Should we expand to the West Coast?" is better than "Let's explore sales data."

2. Data Sources: Document every file, database, API, or system. Include version numbers, extraction dates, and who provided it. This becomes your data lineage foundation.

3. Specific Sub-Questions: Break your core question into 3-5 answerable parts. "What's our West Coast market penetration?" "How do shipping costs compare?" These guide your EDA focus.

4. Success Criteria: What would make this analysis actionable? "Identify three viable expansion cities" gives you a target.

5. Known Constraints: Budget limits, time pressure, missing data fields—write them down so they inform your approach, not surprise you later.

This documentation takes 15-30 minutes but saves hours of unfocused exploration. It creates accountability and ensures your analysis stays aligned with business needs.

Key Takeaway: Document your objectives, data sources, and specific questions before opening your dataset—it transforms EDA from wandering curiosity into focused investigation that answers real questions.