From Data to Graphs to Words - But Where are the Models?


Authors: 
Weldon, K. L.
Pages: 
8-Jan
Year: 
2005
Publisher: 
IASE/ISI Satellite, April
Abstract: 

The pioneers of statistics focused on parametric estimation and summary to communicate statistical findings. The tradition of basing inference on parametric fits is a central mode in statistics education, but in statistics applications, computer-based graphical summary is playing an increasingly important role. A parallel development has been the spread of statistics education to almost all disciplines, and thus the need to communicate statistical results to non-specialists has become more acute. These influences of more graphics and a wider distribution require adaptation in our statistics courses. This paper provides examples of, and arguments for, the use of simulation and graphical display, and the role of these techniques in enhancing the verbalization of analytical results. The immediate goal of the paper is to persuade those who design curricula for early statistics courses to provide a serious introduction to grpahical data analysis, at the expense of some traditional parametric inference. The goals is to enable more students to communicate statistical findings effectively.

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