About this video
Data sets can be split into categorical data (labels or categories) and numerical data (measurements or quantities). Numbers alone do not make data numerical: course numbers like 95 or 101 are really labels. The video then works through a coach recording race times, which is numerical because time is a measurement.
You'll be able to
- tell the difference between categorical and numerical data.
- explain why data that uses numbers can still be categorical.
- decide what type of data a question will produce.
Key words
- data set
- A collection of information, like a list of grades or rainfall amounts.
- categorical data
- Data that puts things into groups or labels, like letter grades or names.
- numerical data
- Data made of numbers that measure or count something, like test scores or rainfall in millimetres.
- measurement
- A number that tells how much of something there is, like a time or a length.
Transcript
- [Instructor] In this video, we're gonna think about types of datasets. And when we think about datasets, broadly speaking, we can divide them into categorical, categorical and numerical datasets. Now, categorical datasets are ones where we're putting things into, as the name implies, categories.
We're putting labels on them. They're saying something meaningful, but they're not necessarily measurements, things that are being quantified. So for example, if I wanted to say, let me categorize the names in a classroom, names in a classroom. Who has what name or even what grades someone has.
So if I said, okay, the grades in this classroom are A, another A, a B, a C, or a D, these are all categories. These aren't quantified things. Now you could imagine numerical involves numbers, but not always involves numbers. For example, if I were to say the grades in the class, and I would say one student got a 97, another student got a 97, another student got an 85, another student got a 73.
This is definitely numerical data right over here. This is quantified. I can manipulate it. It's actually measuring something. Similarly, if I were to measure the rainfall over several days in millimeters, and I were to get 3.20, 0.0, 8.5, 1.7, so on and so forth, that is also numerical data.
Now, on the other hand, if I were to say, what course are you taking? And let's say the course has numbers. So these courses don't have names, and there's course, one student is taking course 95, and the next student is taking course, let's say 101. So these are classes that maybe someone's taking at a university, even though it involves numbers over here, these really are labels or categories.
So this isn't numerical data, this is categorical data. So let's do an example. So here we are told that a coach records the time it takes each runner to complete a race. What type of dataset does this question produce? Pause the video. Try to think about that.
Well, before I even look at the choices, the fact that we're recording the time it takes, these are measurements. So this is going to be numerical. Numerical. So let's look at this choice right over here. The data's numerical because we can count the number of runners.
Now that's not what we're actually looking at. We're looking at the time it takes each runner to complete a race. It is numerical. The dataset is categorical because the runners can be labeled as fast or slow. Well, if they were labeled as fast or slow, that fast or slow list would be categorical data.
But that's not what's going on here. So that's not what we're talking about. The data is numerical because we measure time with numbers. Yes, and the key is it's not just involving numbers, it's that it's an actual measurement. The numbers mean something. The data's categorical because there are many runners in the race.
No, that doesn't really make much sense.
Video © Khan Academy, used under CC BY-NC-SA. Watch on Khan Academy. Puzzicle is not affiliated with Khan Academy.