# Types of data

#### Objective

At the end of the lecture and having completed the exercises students should be able to:

- Describe the different types on variables and give examples of each relevant to dentistry

The methods of description and analysis we apply to data sets are dependent on the type of variable we are considering.

### Categorical data

Variables of this type have values that can be distinguished from each another.

#### Nominal

Nominal variables have values that can be distinguished from each another. An example might be the variable 'Sex' which can have values 'Male' and 'Female'

#### Ordinal

Ordinal variables not only has values that can be put into some sort of meaningful order. If we asked patients how much pain they felt and gave them the options of 'None', 'A bit', 'Quite a lot', 'More than I can bear' then we could put their responses into a meaningful order. We could say that 'Pain' was an ordinal variable.

### Metric data

*Metric* variables can not only have their values put into a meaningful order but can be measured on an 'equal interval' scale. For a metric variable the difference between *1* and *2* is the same as the difference between *2* and *3*; this is *not* the case for an *ordinal* variable.

#### Discrete

Discrete variables are countable variables such as the number of teeth a patient has.

#### Continuous

Continuous variables can be measured to any degree of accuracy, in theory. Blood pressure is a continuous variable.

### Summary

It is important to be able to distinguish different types of data from one another as we use different techniques to describe and analyses the different types. The distinction between *categorical* and *metric* data is more important the the subdivisions of those types. There are other distinctions which are not so important for us at this point. Perhaps the most common of these is the distinction between *ratio* and *interval* subdivisions of *continuous metric* data.

### Another type

*Ratio* class data is continuous data which has a **true zero**. For example *weight* in kilos is a *ratio* class variable: someone who weighs 80kg is *twice* as heavy as someone who weighs 40kg. (I.e., the values can be expressed properly as a ratio.) *Temperature* measured in degrees Celcius, however, is merely an *interval* class variable. Something that is at 20°C is **not** twice as hot as something at 10°C. The two values cannot be properly expressed as a ratio as the zero-point on the Celcius scale is arbitrarliy chosen (at the freezing point of water). Do not worry too much about this distinction as it is not necessary for this course.