PS296 Chapter Notes - Chapter 9: Dependent And Independent Variables
Factors That Affect the Correlation:
Range restrictions:
-the effect of range restrictions of X and Y alters the orrelation from X and Y from what it
would be if the range had not been so restricted
-correlation may either rise of fall as a result, though mostly r is reduced
-restricting the range of X will increase r only when the resitriction results in eliminating
some curvilinear relationship
Example: reading and age data; instead of making range of age (X) 0-70 which would result
in curvilinear relationship as skills decline in old age, resitricting the range of X to 0-17
would result in a higher correlation
Heterogenous Subsamples:
-data in which sample of observations can be subdivided into two distinct sets on basis of
some other variable
-the relationship b/w two variables may be obscured or enhanced by the presence of a third
variable
-ex comparing males and females on the basis of height vs weight
Extreme Observations:
-though something may not seem unusually extreme one either variable X and Y alone, the
combination of the two may be extreme
Document Summary
The effect of range restrictions of x and y alters the orrelation from x and y from what it would be if the range had not been so restricted. Correlation may either rise of fall as a result, though mostly r is reduced. Restricting the range of x will increase r only when the resitriction results in eliminating some curvilinear relationship. Example: reading and age data; instead of making range of age (x) 0-70 which would result in curvilinear relationship as skills decline in old age, resitricting the range of x to 0-17 would result in a higher correlation. Data in which sample of observations can be subdivided into two distinct sets on basis of some other variable. The relationship b/w two variables may be obscured or enhanced by the presence of a third variable. Ex comparing males and females on the basis of height vs weight.