Teaching statistical regression: Simplifying a principle that is hard to understand


Authors: 
McMillan, J. H.
Category: 
Pages: 
3-Jan
Year: 
1989
Publisher: 
Annual Meeting of the American Educational Research Association, March
Place: 
San Francisco
Abstract: 

Statistical regression to the mean is not an easy concept to grasp, especially by students whose background in statistics and probability is limited. As a source of internal invalidity, however, it is an important concept to understand. It is perplexing to students to simply indicate that extreme scores regress to the mean; it is beyond most students to use a more complex mathematical approach. The procedure described in this paper is sufficiently detailed to show students how regression occurs without presenting complicated math. The procedure is based on earlier suggestions by Cutter and Levin.

The CAUSE Research Group is supported in part by a member initiative grant from the American Statistical Association’s Section on Statistics and Data Science Education

register