Track 2 · Exploratory Data Analysis

The core craft of data science. Before any model, any dashboard, any decision — someone has to understand the data. That someone is about to be you.

What you’ll be able to do

• Load tabular data with pandas and inspect its shape, types and quality
• Handle missing values defensibly
• Summarise with describe(), value_counts() and groupby()
• Visualise distributions and relationships with matplotlib
• Compute and correctly interpret correlations
• Measure how certain a group’s number is before it becomes a headline — and handle very small groups responsibly

The dataset

You’ll work with EU Media Habits 2026 — a survey of 624 respondents in 8 European countries about how they consume news and media: hours of TV, social media and news per day, their main news source, and how much they trust the news. (Synthetic data, generated for teaching — but it behaves like the real thing, missing values included.)

Lessons

#	Lesson	Time
1	Explore a real dataset	~90 min
2	Graded exercise — with instant feedback	~60 min

Complete both and you’ll have done a genuine end-to-end exploratory analysis — the same workflow used on real projects, just on a friendlier dataset.