CHAPTER 01 · TOPIC 02
Types of Data
Distinguish categorical and numerical data and the nominal, ordinal, interval, and ratio scales.
Before choosing a statistical method, identify the type of data. Data type determines which operations are meaningful and which summaries, graphics, and analyses are appropriate.
Categorical data#
Categorical, or qualitative, data classify observations by a characteristic. Their values represent group membership rather than quantities that can be added or divided.
- Nominal scale: categories have no inherent order, such as blood type, nationality, species, or treatment arm
- Ordinal scale: categories have an order, but gaps between adjacent levels need not be equal, such as pain severity, disease stage, or satisfaction
Numerical data#
Numerical, or quantitative, data are counts or measurements for which magnitude and differences carry substantive meaning.
- Discrete data arise from counting and usually take separated values, such as admissions, cases, or number of children
- Continuous data arise from measurement and can in principle take arbitrarily fine values within a range, such as height, weight, blood pressure, or temperature
Levels of measurement#
| Scale | Order | Equal differences | Meaningful zero | Examples |
| Nominal | No | No | Not applicable | Blood type, nationality |
| Ordinal | Yes | Not necessarily | Not applicable | Pain grade, disease stage |
| Interval | Yes | Yes | Relative zero | Temperature in °C |
| Ratio | Yes | Yes | Absolute zero | Height, weight, elapsed time |
Why data type matters#
Categorical variables are often summarized with counts and proportions. Continuous variables may be summarized with a mean and standard deviation or a median and interquartile range, depending on their distribution and the purpose of the analysis.