The Centre of the Data and the Effects of Extreme Values

The mean and median are familiar ways to summarize the values of a numerical variable.  In many situations the values of the mean and median are similar, but unusual observations can lead to quite different results.

Stream the video without the embedded quiz questions by clicking on the video link below. Closed captions are available.

 

Notes on the video: The Centre of the Data and the Effects of Extreme Values

Some points to consider for this video:

  • How does a modified box plot illustrate extreme values?
  • What does it mean for a measure of centre to be robust?
  • There are no set rules for what percentage of the data should be trimmed in a trimmed mean. The goal is to remove the extreme observations. If fact, you can trim as much as 50%. What is another name for the 50% trimmed mean?