Marie said:  Now what doesn't make sense to me that two groups individually 
have positive and significant correlations but the two groups combined can have 
a negative and significant correlation.
Here is an example I used when teaching Methods to students in a Criminal 
Justice program. Imagine two groups: psychology students and police officers. 
In both groups, you measure knowledge of psychology and skill at interrogating 
suspects. It is quite reasonable that, within each group, those who know more 
psychology will be better at interrogating suspects, so within each group these 
two variables have a positive relationship. However, when looking overall, we 
see one cluster high in psych knowledge and low in interrogation skills, and 
another cluster low in psych knowledge and high in interrogation skills, so the 
group as a whole shows a negative correlation.

Celia
*  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *  *
Celia Reaves, Ph.D.
Department of Psychology
Monroe Community College
1000 East Henrietta Road
Rochester, NY  14623
Phone:  585-292-3258
email: [email protected]<mailto:[email protected]>

From: Helweg-Larsen, Marie [mailto:[email protected]]
Sent: Monday, February 27, 2012 12:08 PM
To: Teaching in the Psychological Sciences (TIPS)
Subject: RE:[tips] Statistical question-correlations


---
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=16300
or send a blank email to 
leave-16300-13090.68da6e6e5325aa33287ff385b70df...@fsulist.frostburg.edu

Reply via email to