Different data require different methods of summarising, describing and analysing.
There are four main types of data: Nominal, Ordinal, Interval and Ratio. It is important to be able to identify which type of data you have in order to choose appropriate statistical methods. Take a look at the examples below for a quick refresher of the four types of data:
Nominal data are named variables. Nominal data is unordered, categorical and mutually exclusive - which means that each category is separate and cannot occur at the same time.
Examples of nominal data are:
- Different countries: Afghanistan, Brazil, China, etc.
- Yes or No answers. When a variable only takes only two values, we call this variable dichotomous.
Ordinal data are named variables that have a meaningful order. Ordinal data is ordered, categorical and mutually exclusive (cannot happen at the same time).
Examples of ordinal data are:
- Body Mass Index (BMI): Underweight, normal weight, overweight, obese.
- Responses to questionnaires: Strongly Disagree, Disagree, Neutral, Agree and Strongly Agree. This is known as a Likert scale, and is often coded using numbers where:
1 = Strongly Disagree
2 = Disagree
3 = Neutral
4 = Agree
5 = Strongly Agree.
How should I analyse Likert data? Take a look at the following article Analysing LIkert Scale/Type Data from Maths Support at St Andrews University for some ideas.
Interval data is quantitative (numbered data) that has a meaningful interval between data points but does not have a meaningful zero, where zero means nothing.
Examples of interval data are:
- Temperature (in degrees Celsius, 0 degrees Celsius is cold, not no temperature)
- Size of shoes (EU size 0? This shoe size doesn’t exist.)
Ratio data is quantitative (numbered data) that has a meaningful interval between data points. Unlike Interval data, ratio data has a meaningful zero - known as a true zero.
Examples of ratio data are:
- Length (in cm, where 0 cm has no length).
- Weight (in kg, where 0 kg has no weight).