Data comes in various forms and it is important to properly classify different types of data in order to analyse and extract meaningful insights from it effectively. There are four main categories that data can typically fall under – nominal, ordinal, discrete and continuous. Proper identification and distinction between these four types of data is crucial for statistical analysis and making data-driven decisions. In this blog, we will discuss these four types of data in detail, including the differences between them.
Nominal and Ordinal Data
Qualitative data describes by categorizing characteristics and attributes without specific numbers attached. It groups similar features together.
Nominal Data
Nominal data, also known as categorical data, assigns objects or variables into categories without any numerical value or order. The categorical labels themselves do not reflect any ranking or hierarchy among the categories.
For example, gender (Male, Female), colour (Red, Blue, Green), country of origin, etc., are all examples of nominal data since they categorise objects into labels without any inherent ordering. Names of each category represent nominal data as pie charts or bar graphs.
Ordinal Data
Ordinal data differs from nominal in the sense that while it also assigns categories, there exists a logical order or ranking among the categories. However, the ranking is subjective and does not denote the exact magnitude or difference between categories.
For instance, popularity/rating of movies (Blockbuster, Hit, Flop) and performance evaluation ratings (Exceeds Expectations, Meets Expectations, Needs Improvement) are all types of ordinal data since the categories can be logically ranked or ordered. Ordinal data is usually represented through bar or line graphs.
The key difference between nominal and ordinal data is that while nominal data categories do not imply any ordering or ranking, ordinal data categories do follow a hierarchy or ranking according to some attribute, even if the scale of difference is not equitably measured.
Discrete and Continuous Data
Discrete and continuous data fall under the umbrella of quantitative data. Quantitative data refers to numerical attributes that can be measured, counted or continuously variable.
Discrete Data
Discrete data consists of separate, distinct, discontinuous values or points that can be counted. Discrete variables take on integer values, such as the number of children in a family or the number of books sold by an author.
Some key properties of discrete data include:
- Can be counted but not measured continuously
- Represented through bar graphs as individual data points are separated
- Examples include a number of customers, light bulbs burnt out, coins in your pocket, etc.
Continuous Data
Continuous data consists of measurements that can take any real number value within a given range. Unlike discrete data, continuous variables can be measured precisely and may have an infinite set of possible values within a fixed interval.
Key aspects of continuous data include:
- Can be measured precisely as a real number
- Represented through line graphs or histograms
- Examples include temperature, time, weight, height, test scores on a scale, etc.
The main difference between discrete vs continuous data is that discrete can only take on integer values, whereas continuous can be measured precisely as real numbers within a range.
Nominal vs Ordinal vs Discrete vs Continuous
To summarise the key differences between these four data types:
- Nominal data categorises objects without implying ordering. Ordinal does imply ordering but not scale.
- Discrete variables take on integer values and points; continuous are real numbers within ranges.
- Nominal and ordinal are qualitative, while discrete and continuous are quantitative.
- Nominal is shown by names, ordinal by rank. Discrete by separate points, continuous as smooth curves.
The proper identification and distinction between these four types is important for determining appropriate statistical tests and analysis techniques on any given dataset.
Conclusion
Understanding how to distinguish between and classify different data is a crucial starting point for any data science project. It guides subsequent steps like wrangling, cleaning, modelling and interpreting results. With a firm grasp of data types, analysts can efficiently analyse any dataset.
