Data Types

Practically speaking, data can be divided into categorical variables and numerical variables. The type of variables you measure determines the type of statistical analysis that is appropriate

Categorical Variables

Categorical variables are selected from established categories. Of note, the categories must be clearly defined. Examples include gender (male vs. female), pregnancy (pregnant vs. not pregnant)

Numerical Variables

There are three different kinds of Numerical variables:

  • Ordinal variables are numerical, non-numerical, or groups of variables which belong in a meaningful order. For example, you could define a patient’s blood sugar as being hypoglycemic if less than 80, normoglycemic if between 80 and 120, and hyperglycemic if greater than 120. Categorizing each continuous piece of data into one of the three categories creates ordinal data. Another example of an ordinal variable is the BIS score – while a BIS of 41 is higher than a BIS of 39 (and reflects a meaningful increase in the likelihood of awareness), there is no such thing as one unit of “BIS.” Pain scores are a third example of ordinal variables – 10/10 pain is higher than 9/10 pain but there is no underlying mathematical meaning to one point on the numerical pain score
  • Discrete variables are counts of things. Examples include the number of times a patient vomited in the PACU
  • Continuous variables are measures where, in theory, and number may occur (or any number within a range of values may occur). Examples include hemoglobin, blood pressure, and temperature

Retrieved from “