Our third measure of central tendency is the mode. The mode is just the value that appears the most often in the dataset. If there's no unique value that appears the most often (for example, maybe there are two or more values that appear the most often), then the dataset has no mode.
The mode applies to both categorical and discrete numerical data, though for numerical data the mean or median are usually better measures of central tendency for. For categorical variables the mean and median aren't available to us. So we use the mode for categorical variables.
Here are some examples.
Example: The mode of numerical data
Consider the numerical dataset
3, 8, 2, 8, 4, 8
The mode is 8, since that appears more than any other individual element in the dataset.
Example: The mode of categorical data
Consider the following categorical data:
INFO, INFO, TRACE, TRACE, TRACE, INFO, DEBUG, TRACE, TRACE, ERROR, TRACE
Here the mode is TRACE, since that appears more times (6) than any other element in the dataset.
Example: Undefined mode
Finally, consider the following dataset:
BIDEN, BIDEN, TRUMP, TRUMP, BIDEN, TRUMP, YANG
Here, both BIDEN and TRUMP both appear three times, and YANG appears only once. There is no mode, because there is no unique value that appears more than all the other values.
Exercise 1. What is the mode of the following dataset?
AMZN, MSFT, MSFT, AMZN, GOOG, AMZN, SBUX, UBER, AMZN
Exercise 2. In what sense is the mode a measure of central tendency?