Improving statistics education through simulations: the case of the sampling distribution


Authors: 
Earley, M. A.
Category: 
Year: 
2001
Publisher: 
Paper presented at the Annual Meeting of the Mid-Western Educaitonal Resaerch Association, Chicago, Illinois, October 24-27, 2001
Abstract: 

This paper presents a summary of action research I am doing investigating statistics students' understandings of the sampling distribution of the mean. With 4 sections of an introductory Statistics in Education course (n=98 students), I have implemented and evaluated a computer simulation activity (delMas, Garfield, & Chance, 1999) to show students the sampling distribution "in action," as recommended by many authors and statistics professors. Assessments after the activity point to clear deficiencies in student understanding, such as not understanding how sample size affects the shape and variability of sampling distributions. With the addition of the computer simulation, students' understandings are still incomplete. This is most evident as I move forward in class and introduce inferential statistics. My results discuss some of the deficiencies I have identified in student understandings. Also included is a discussion of how the action research cycle will continue to work on these deficiencies.

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