The Data Verse - 01
We are living in times when every other conversation leads to something involving artificial intelligence. It has become a digital assistant of sorts, helping us with everyday tasks like predicting the next word we are going to type, finding the fastest route, recommending what to watch or buy in streaming services and shopping platforms, or breaking down complex tasks into simpler ones. There us always a new tool, model, or algorithm claiming to be better and promising to improve the way we work. And somewhere between the fancy names and technical jargon, a fundamental component is overlooked.
The data itself.
Data scientists, analysts, and other professionals spend hours learning algorithms, comparing models, and trying to achieve the best possible performance for the problem at hand. But somewhere in this process, are we overlooking a much simpler question: Have we actually understood the data we're working with?
Consider something as simple as a spreadsheet containing student details. It can have fields like student IDs, ages, attendance, grades, ratings, and written comments. The student ID might look like a number, but taking the average of student IDs would be meaningless. A feedback rating of 4 might be higher than 2, but can we say that it’s exactly twice as satisfactory? What about the written comments? How do we compare a sentence with a percentage?
This is where things get interesting. Understanding data isn’t just about identifying whether it is a number or a non-number, nor is it about just knowing the statistical concepts. It's about understanding what these values represent, and we can leverage them to draw meaningful conclusions.
I do believe that every dataset has a story to tell. But are we looking at it closely to understand it? Probably that’s where our Data Verse journey should begin.
For a detailed explanation of types of data, you can explore this article from Turing , which explains how data can be categorized based on its nature and level of measurement. The distinctions are not new, especially for learners and professionals familiar with statistics. However, understanding what they mean is crucial when we start working on tasks dependent on the data. Remember the student dataset we discussed earlier? Let us see how these classifications can be applied to it.
| Variable | Type of data | Description |
| Student ID | Nominal | Although represented by numbers, it is a unique identifier and cannot be considered quantifiable |
| Age | Quantitative – Ratio | A measurable quantity with a meaningful zero |
| Department | Qualitative – Nominal | A category without any natural order |
| Grades (O, A+, A, B, C, D) | Qualitative – Ordinal | Categories with a meaningful ranking |
| Number of classes missed | Quantitative – Discrete | Count of classes missed, and it’s a whole number |
| Height | Quantitative – Continuous | A measurement that can take values within a range |
| Attendance Percentage | Quantitative – Ratio | Attended classes expressed as a percentage |
| Written Comments | Qualitative – Textual | Descriptive information expressed in the form of text instead of numbers |
| Ratings (1-5 or Poor-Excellent) | Qualitative – Ordinal | Levels of student satisfaction ratings that are ordered, but differences between ratings are not necessarily equal. |
Now, How About Trying It Yourself
Pick a dataset from any public repositories such as Kaggle or the UCI Machine Learning Repository, and try classifying the data based on what we have discussed. It doesn't have to be a large or complicated dataset. Even a small one with just a few columns is a good place to start.
And here's something to think about. How would you classify a date? Sounds simple, doesn't it? But give it some thought. It can be a little tricky!
If you would like to share it, upload your work and drop the link in the comments section. I would love to see your approach. Sharing isn't mandatory, though. But do give it a try! You might be surprised at how much you notice when you start looking at data a little more closely.
That’s all for now. See you soon with the next article of The Data Verse series. Keep exploring till then!
Leave a Reply