What Is Data Analysis?

Questions, data and answers

Every organisation collects figures as it goes about its work. A tea house records each item it sells; a school records every pupil's marks; a weather office records the rain that falls each month. Most of the time these figures simply pile up. Data analysis is the work of turning them into useful answers:

• Which item on the menu sells best, and which hardly sells at all?
• Are sales rising or falling, and do they change with the monsoon?
• Which subject do pupils find hardest, and has it improved since last term?
• Is this month's rainfall unusual, or quite normal for the time of year?

Data analysis always follows the same three steps. First you ask a clear question. Then you look at the data: you choose the rows and columns that matter, tidy up any mistakes, and calculate totals, averages or comparisons. Finally you give an answer, often with a chart, that someone can act on. The calculations are the easy part once you know how; asking good questions and checking that the answer makes sense are what make a good analyst.

Data comes in tables

Most data is kept in tables, just like a page of a spreadsheet. Here are the first few rows of the practice data you will use most in this course: the sales of The Malabar Tea House, a small tea room in Calicut with two branches, for the whole of 2025.

Each row is one item on one customer's order: order 1001 was one Masala Chai and three Cardamom Milk Teas, paid in cash. Each column holds one kind of information for every row. The whole year has 7,202 rows: far too many to add up by hand, but no trouble at all for a computer.

Why Python and pandas?

Spreadsheet programs are fine for small tables, and many people start there. Analysts turn to a programming language when the data grows, when the same analysis must be repeated every week or month, or when they need to show exactly how an answer was reached. This course uses Python, the most widely used language for data analysis, together with two free tools written for it:

• pandas, which holds a table in memory and lets you select, filter, group and summarise it with short, readable instructions;
• matplotlib, which draws charts.

You do not need to have taken the Python course. Lesson 1.3 covers the small amount of Python you need, and everything else is explained as it comes up.

Where do AI tools fit in?

You can ask an AI tool to “analyse this spreadsheet”, and it will often write pandas code very like the code in this course. That is useful, but the tool does not know your business, it cannot tell whether the data has mistakes in it, and it sometimes answers a slightly different question from the one you meant. Someone still has to ask the right question, read the code, and check that the answer is sensible. Module 9 is devoted to exactly that: averages that mislead, patterns that are not causes, and how to check results, including results written by an AI.

How this course works

This is a text course with no videos, but it is not a course you only read. Every lesson has boxes of real Python code that you can run and change, right here on the page.

Try it now: tap the blue Run button below the box. The first time you do this, your browser downloads Python and the data tools (about 30 MB). This takes a little while, longer on mobile data, and it happens only once: after that your browser keeps them, and every box runs in a moment.

In a few lines, Python read the whole year of sales, worked out the money taken for each row, and added it up for each category: cakes, coffee, snacks and tea. You will learn exactly how each line works over the next few lessons. For now, try changing "category" to "branch" (keep the quotes) and press Run again: you will see the takings of each branch instead.

Charts are just as easy. This box draws the takings of each month as a line:

Can you see the dip in the middle of the year, and the rise at the end? A chart like this raises questions straight away: is the dip the monsoon? Is the December rise the festive season? Asking and answering questions like these is what this course is about.

What you will learn

Module 1 gets you started with tables in pandas. Modules 2 to 5 teach you to explore, select, clean and summarise a table. Module 6 deals with dates and trends, Module 7 with combining tables, and Module 8 with clear, honest charts. Module 9 is about thinking like an analyst, and in the final project you analyse the tea house's whole year and write up what you find.

Quick check

Three questions to confirm the main ideas before you move on.

Found a mistake on this page, or something unclear? Report a problem and mention “Data Lesson 1.1”.