What Type is That Data and Why Does It Matter Anyway?

One of the first things I learned as I began to explore data analysis was that data can be categorized into different types. It can be either categorical or numerical, and each of these types can be further divided into two categories.

Categorical data is also called qualitative. It can be nominal or ordinal. Both are ways of putting data into categories, but the categories for ordinal data also have an order associated with them. Examples of nominal data are ‘cat’, ‘dog’, ‘bird’, and ‘hamster’. There is no implied order in these types of animals. While some people might prefer one type of pet over another, there isn’t a generally agreed-upon standard for which pets are better than others. On the other hand, the ordinal categories ‘poor’, ‘fair’, ‘good’, and ‘excellent’ do indicate an order. ‘Fair’ is better than ‘poor’, ‘good’ is better than ‘fair’ and ‘poor’ and ‘excellent’ is the best of them all.

Numerical data is also called Quantitative. It can be interval or ratio. The main difference between these two is that interval data doesn’t have a ‘true zero’. This means that either zero isn’t a possible value or a zero on the interval scale doesn’t mean that there is absolutely nothing of what is being measured. For example, someone can’t have an IQ of zero, and at 00:00:00 in the morning all time hasn’t ceased to exist.

Understanding the different types of data wasn’t difficult, nor was using this information to determine which types were in the data I was working with. However, I only actually appreciated how this information could be useful after I came across guidelines for using each type of data in a book. The book explained what types of analysis could be used with each kind of data. It also listed mathematical operations that could be used and transformations that could validly be made for each type. I had understood some of these rules intuitively. For example, I never tried to find the average of any categorical data I worked with. (What would the average of 5 cats, 7 dogs, and 3 birds be anyway?) Even so, I had not been using this information explicitly while planning how I would use data to answer questions. Now that I have this insight, I will use it when planning future analysis.