In the last article, we learned about what are variables. Now in this, we are going to learn about the variable measurement scales. In this you might know a few terms if you are from a math background and you might not know some concepts so be sure to read the entire article.
In Variables, there are 4 major types of Measurement Scales
4 Types of Measurement Scales
Nominal
Ordinal
Interval
Ratio
You might be thinking that yes we know all these types of measurement. but it is very important to learn about this because the dataset that you might be working on has this type of data.
The data can be Nominal, Ordinal, Interval and Ratio related.
So if you know about these measurement scales you can do very good kind of data analysis on your dataset
Now, let's explore these types of measurement scales deeply
Nominal Data
Nominal Data - Nominal data is categorical data
eg: Colors - In colors you have different types of colors such as Red, Blue, Black, and Orange.
Another eg of Categorical data is
Flowers - In flowers also you have different types such as Rose, Marigold, lily, Jasmine, Lotus etc.
So this type of data which is categorical data comes under Nominal data.
Ordinal Data
Now like Nominal data, Ordinal data can't be properly defined. So to understand this type we will take an eg
Remember: In Ordinal data, the order of the data matters but the values do not matter.
Student (Marks) | Ranks |
100 | 1 |
96 | 2 |
49 | 4 |
84 | 3 |
35 | 5 |
The Ranks in the above eg can be called Ordinal data
you can see the ranks 1, 2, 4 and 3 so by observing them we can say that Ordinal data focuses more on the order of the data and not the values (marks).
In this case, we will not look at the marks of the student but we will consider their ranks.
Interval Data
In the Interval Data type unlike Ordinal data where value does not matter but order does here Order of the data matters, the values also matter but there's a little catch the natural 0 is not present here.
let's consider an example of temperature here
Fahrenheit |
70 - 80 |
80 - 90 |
90 - 100 |
In the above example, 0 Fahrenheit won't make a useful interval. hence the natural 0 is not present here in the Interval Data.
Ratio Data
Ratio data can be defined as the measured numerical data that have an equal distance between adjacent values and meaningful zero.
Ratio data can be continuous and discrete but ratio data cannot be negative.
let us take an eg to understand the ratio data properly
eg:
Age from 0 - 100+ years
Temperature (in Kelvin, but not in Celcius or Fahrenheit)
Time Interval (measured with a stopwatch)
In this eg, there is a real and meaningful zero point. Temperature data in Celcius or Fahrenheit can be negative so it will not be considered in Ratio Data.