Categorical data consist of labels; numerical data consist of measured or counted values. Discrete values are counted in separate steps, while continuous measurements can take any value in a relevant interval. A grouped frequency table collects values into nonoverlapping classes.
Use a bar chart for separate categories and a histogram for grouped continuous measurements. A scatter graph represents pairs of numerical measurements. A time series keeps time on its horizontal axis. Choose a display for the data type and the question, rather than its decoration.
Check the conditions
Use the stated units and precision.
Show the method as well as the final answer.
Three worked examples
Example 1 · Data types and suitable displays
Consider the data situations described in each question. Categories and measurements have their ordinary statistical meanings.
Classify preferred school subject.
discrete numerical
continuous numerical
a probability
categorical
A table has 7 observations in one category and 5 in another. Find its total frequency.
Of 35 measurements, 7 are excluded. How many remain?
A two-way table has row totals 7 and 15. Find the grand total.
Show the reasoning
(a)
Distinguish labels, integer counts and continuous measurements.
Preferred school subject is categorical.
Answer: categorical.
(b)
Sum each category count once.
Add the two nonoverlapping counts to obtain 12.
Answer: \(12\).
(c)
Distinguish excluded records from retained observations.