  ## Data Measurement Scales and Different Types of Data

As we discussed in the first part of this article – the different types of data, we can divide data into two main categories – qualitative or quantitative (numerical or not )

From there, numerical data can be sub-divided into discrete or continuous data, based on the type of numerical data it consists of.

Now let’s have a look at the next layer of subdivisions, we call these subdivisions the measurement scales of data:

• There are four main types of data measurement scales.

These so-called scales are simply subcategories to the main divisions of different types of data (qualitative and quantitative)

• Continuous Data (a subsection of Quantitative data) can be further divided into Interval and Ratio data types.
• Qualitative data can be further divided into Nominal and Ordinal data types.

We will dissect each scale below, and bring some examples to clarify their meaning.

# Nominal and Ordinal Types of Data Scales:

Nominal and Ordinal Scales are subsections of Qualitative data – therefore, they are not numeric.

## Nominal Data Scales

The Nominal data subcategory is one of the easiest to understand out of all the scales, so let’s start with this one.

Nominal scales are used for labelling variables that don’t have a numeric value. They could technically be called labels.

Multiple-choice questions usually collect this type of data.

Think of the word nominal – it kind of sounds like the word name, so it might help you remember that nominal scales are like labels or titles (or names).

See some examples below:

• Rose
• Tulip
• Daffodil
• Poppy
• Sunflower

• Male
• Female
• Other

Nominal Category can also be divided into nominal with order (like slow, fast, fastest) and nominal without order (male, female)

## Ordinal Data Scales

Ordinal sounds like order – this is the best way to remember this as the scale is all about data, where the order is relevant.

It is similar to Nominal data in terms of it usually being a multiple-choice questionnaire for collecting this type of information. However, in comparison to Nominal data, each of the labels now has an order of some sort.

See the example below:

How was our service?

• Not great
• Average
• Good
• Great

In this example, we can tell that option two is better than option one, and that option four is better than option three. However, we can’t measure that difference between those options; we can’t determine how much better one option is from another.

The classical use for these scales is for customer happiness surveys, as they usually focus on non-numeric concepts like satisfaction and happiness.

# Interval and Ratio Types of Data Scales

Let’s move our focus back to the numbers.

Interval and Ratio Data scales are a subsection of Continuous data, which is a subsection of quantitative (numeric) data.

This means that we are dealing with numbers that can be divided into increments and shifted from one measuring scale to the other.

## Interval Data Scales

Interval Data deals with numbers and order. Interval means space in between, so interval scales tell us about the order, but also the values between them.

Let’s use the temperature as an example.

The difference between 20 degrees Celsius and 30 degrees Celsius is 10 degrees; this is the same as between 40 degrees and 50 degrees. It is a measurable difference.

However, the Ratio scales do not have a zero point. The temperature example is excellent here because there is no such thing as no temperature. There is always a temperature, even when it is 0 degrees Celsius because, in this scenario, zero does not represent a lack of temperature. This scale can also go into the negatives.

The issue with this scale is because it does not have a true zero; it is not possible to compute ratios. We can add and subtract, but not multiply or divide.

Therefore we can’t transfer the data from one measurement system to another.

## Ratio Data Scales

Ratio data scales are the most complex and the most detailed out of all the data scales. They tell us about the order, the values between the units and also have an absolute zero.

The existence of an absolute zero allows for a range of descriptive and inferential statistics to be applied.

Everything about Interval data scales applies to Ratio data scales with the addition of having an absolute zero.

Weight, height, length and duration would be great examples of ratio data.

Because these variables can be added, subtracted, multiplied and divided, ratio data comes with many possibilities for statistical analysis.

# Summarizing the data scales

To sum this up:

• Nominal and Ordinal data scales deal with non-numeric data – they deal with words.
• Nominal is for names or labels for a series of values with no order.
• Ordinal id the same as nominal, but there is an order of some sort involved.
• Interval and Ratio scales deal with numeric data
• Interval deals with an order of values that have a set difference between each one.
• Ratio data is the same as Interval data but has a true zero, so it gives us the ability to calculate ratios.

Who are we?

If your data is out of date and needs cleaning, have a look at our data cleaning service.

DataCleanse is a product of MMR. We are a full-service agency with expertise in design, marketing, eBrochures, digital, etc. If there is anything else we can help you with, feel free to get in touch.

Find out how Datacleanse can work for you.

Contact us on +61 3 9090 7000 or enter you details below and we will be in touch to discuss your custom solution.