Here's the whole idea in one breath: data is just recorded observations. The moment you write something down — a score, a colour, a temperature — you've made data. It's not a special substance that lives in computers. It's the world, noticed and noted.
It's Already Everywhere
Your phone counts your steps. The footy app tallies goals. The weather bot logs the temperature every hour. None of that is exotic — it's all just somebody (or something) watching the world and recording what they see. Once you start looking, you can't stop: shoe sizes, bus arrival times, how many kids in the class have a dog. All data. The trick of this whole topic is learning to read the kind of data you're holding, because the kind decides what you're allowed to do with it.
The Two Big Families
Every value you ever record lands in one of two families. Categorical data is labels or groups: eye colour, favourite sport, a yes/no answer. You're sorting things into named buckets, not measuring them. Numerical data is numbers you genuinely count or measure: how many goals, how tall, how hot. The quick test — could you sensibly do maths on it? You can average heights; you can't average "blue, brown, green."
Numerical Splits Again
The numerical family has two children. Discrete data is counted in whole steps — number of pets, goals scored, siblings. You can have 0, 1 or 2 dogs, never 1.6 of a dog; it jumps. Continuous data is measured on a smooth scale — height, time, temperature — where between any two values there's always another. You can be 152 cm, or 152.3, or 152.34; it never jumps, you just stop measuring at some point.
So the whole map is: data → categorical or numerical → and if numerical, discrete or continuous. Drag a few values through the toy and the shape of it will click — same world, just sorted into the boxes that tell you what it's safe to do next.