|
Activity Number:
|
509
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Education
|
| Abstract - #305007 |
|
Title:
|
Confound Those Speculative Statistics
|
|
Author(s):
|
Milo Schield*+
|
|
Companies:
|
W.M. Keck Statistical Literacy Project
|
|
Address:
|
9 Robb Farm Road, North Oaks, MN, 55127,
|
|
Keywords:
|
statistical deaths ; observational studies ; statistical literacy
|
|
Abstract:
|
Speculative statistics are model-based statistics. These include deaths attributable to being in a group (deaths linked to a distant cause). Such deaths are those due to primary smoke, second-hand smoke, obesity and radon. This paper reviews the epidemiological model used and introduces a graphical technique to present three big ideas: that a confounder can influence (1) an association in an observational study, (2) the speculative statistics generated by epidemiological models and (3) the statistical significance calculated in comparing these statistics. This paper argues that if students are to deal with the statistics of everyday life, they must appreciate these three big ideas. They must be aware that speculative statistics are often indistinguishable from actual statistics and are vulnerable to confounding. A separate statistical literacy course based on these ideas is recommended.
|
