A Singapore Government Agency Website

Official website links end with .gov.sg
Secure websites use HTTPS
Look for a lock () or https:// as an added precaution. Share sensitive information only on official, secure websites.
Sign In

Pitfalls

The mode is not a very useful average when the distribution is very flat.

This is to say, when the frequencies for all groups are very alike, the mode does not provide much meaningful information. Consider Example 2.

Example 2: Multiple Modes in a Flat Distribution
 
Suppose we asked each of the 20 employees in Company ABC for the training hours completed in the year.
 
Suppose all 20 employees gave different answers ranging from 7 hours to 26 hours with one employee completing each number of training hours, meaning only one person completed 7 hours, only one person 8 hours and so on.  We call such a distribution flat (see following chart) since the number of observations in each category are equal or almost the same. By the definition of mode, any number of hours between 7 and 26 hours is a mode. Therefore, the mode only tells us that each of these training hours were equally likely.
 
Distribution of Employee Training Hours
 

 


The mode can occur at the tail end of a distribution...

…and is not necessarily at or near the middle of a distribution.  At times, there can even be multiple modes. Example 3 explains.
 
Example 3: Modes at the Extreme Ends of the Distribution
 
Suppose we now ask each of the 20 employees in another company, XYZ, for their training hours. For this sample, we can observe two points with the highest number of observations: 8 hours and 28 hours (see following chart). It also illustrates that the mode does not necessarily occur near the centre of the distribution. In fact, like shown here, it can be found at either or both of the two tail ends of a distribution.
 
Distribution of Training Hours of Employees of Company XYZ