Hi all,
Chris Green sent a link to an example of a spurious correlation between rises
in autism rates and sales of organic foods. I've rarely seen such high
correlations in real-world data, so I felt compelled to check the reports to
make sure the data used to calculate the correlation coefficient were accurate.
It took a little time to find all the data used in the graph (the organic-food
data were in several places online), but the correlation of 0.997 is correct!
Oddly, the organic-food data changed slightly between the 2009 and 2011 surveys
(actually, the change first appeared in the 2010 survey data). Beginning in
2004, the amount increased slightly:
2009 Data 2011 Data
2002 8,635 8,635
2003 10,381 10,381
2004 11,902 12,002
2005 13,831 14,223
2006 16,718 17,221
2007 19,807 20,410
I didn't feel like searching through the summary reports to see if there was an
explanation. The only effect on the correlation coefficient was to make it
negligibly higher (0.998).
Best,
Jeff
--
---------------------------------------------------------------------------------
Jeffry Ricker, Ph.D.
Professor of Psychology
PSY 101 Website: http://sccpsy101.wordpress.com/
Knowing Ourselves: http://psysci.com/
---------------------------------------------------------------------------------
Scottsdale Community College
9000 E. Chaparral Road
Scottsdale, AZ 85256-2626
Office: SB-123
Phone: (480) 423-6213
Fax: (480) 423-6298
---
You are currently subscribed to tips as: [email protected].
To unsubscribe click here:
http://fsulist.frostburg.edu/u?id=13090.68da6e6e5325aa33287ff385b70df5d5&n=T&l=tips&o=22629
or send a blank email to
leave-22629-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu
