What Is a Dataset, and What Question Are You Asking?

Understand rows, columns, and how to turn a vague curiosity into a answerable question.

What you will learn

Describe a dataset as rows and columns
Identify what a single row represents in a dataset
Turn a vague curiosity into a specific, answerable question
Recognize the difference between a question and a metric

The idea

A dataset is just a table: rows and columns. Each row is one 'thing' (a customer, an order, a day), and each column is one fact about that thing (their name, the amount, the temperature). Data analysis is the skill of asking a real question, then using the rows and columns to answer it with evidence instead of guessing.

Why does this matter?

Every analytics job, dashboard, and data science project starts here. If you cannot state your question clearly, you cannot know which columns matter, which chart to draw, or when you are actually done. Most bad analysis comes from a fuzzy question, not bad math.

A real-world analogy

Think of a dataset like a class attendance sheet: one row per student per day, columns for name, date, and present/absent. 'Is attendance a problem?' is vague. 'Which weekday has the lowest attendance rate this term?' is a question you can actually answer by looking at the sheet.

See it in code

A tiny dataset: one row per order

const orders = [
  { id: 1, customer: 'Ana', amount: 42.50, city: 'Lisbon' },
  { id: 2, customer: 'Ben', amount: 15.00, city: 'Porto' },
  { id: 3, customer: 'Ana', amount: 8.20,  city: 'Lisbon' },
];

// Question: "What is the total amount spent per city?"
const total = orders.reduce((sum, row) => sum + row.amount, 0);
console.log(total); // 65.7

Line by line

const orders = [ {...}, {...}, {...} ];
This is our dataset: an array of row objects. Three rows, one per order.
{ id: 1, customer: 'Ana', amount: 42.50, city: 'Lisbon' }
Each key (id, customer, amount, city) is a column. This one row describes a single order.
orders.reduce((sum, row) => sum + row.amount, 0)
reduce walks every row and accumulates a running total from the amount column — a first taste of turning rows into an answer.

What do you think happens?

In the orders dataset above, what does a single row represent?

Have a guess before you read on. Guessing wrong is part of learning it.

Try it yourself

Rewrite the vague question 'Is our shop doing well?' as a specific, answerable question using only columns that exist in the orders dataset (id, customer, amount, city).

Worth knowing

Common mistake: confusing a metric ('total revenue') with a question ('is total revenue higher this month than last month, and why?'). A metric is a number; a question tells you what comparison or pattern you are actually looking for.

The proper words for it

Dataset: A structured collection of data, usually organized as rows and columns.
Row (record): One entry in a dataset — one instance of the thing being described.
Column (field): One attribute or fact recorded about every row, like 'amount' or 'city'.
Analysis question: A specific, answerable question that tells you what to measure, compare, or explain.

Where you'll meet this

A support team doesn't just ask 'are customers happy?' — they ask 'what percentage of tickets this month were resolved within 24 hours, and how does that compare to last month?' That precision is what makes the analysis useful.

Lesson recap

A dataset is rows (things) and columns (facts about those things)
Good analysis starts with a specific, answerable question, not a vague feeling
A metric is a number; a question tells you what to compare or explain
You will spend this course turning questions into code that answers them