Friday, April 29, 2011

"two Tailed vs single Tailed"

In doing statistical analysis especially in t test, commonly students are getting confused whether using one-tailed or two-tailed test. Thus, here are some brief distinction between them and wish it could help you to deal with. 
one-Tailed:
"Defined-Prediction" 
1. Researchers expect that experiment procedure has an influence to clear direction (can be defined). 
2. Significance test with one-tailed test must be chosen.
e.g.,
There is no positive influence between intellectual rate of kids and nutrition intake while in period preschool.  
There is positive influence between intellectual rate of kids and nutrition intake while in period of preschool.   

two-Tailed:
"Non-defined Prediction"
Researchers expect that experiment procedure has an influence yet the direction is not clear.
e.g., 
There is no influence between sequence of birth and motivation for achievement in kids.
There is influence between sequence of birth and motivation for achievement in kids.     

CONCLUSION: 
1. two-tailed is more conservative than one-tailed test.
2. two-tailed test is widely applied in scientific research rather than one-tailed.

source: Aron, Aron, & Coups.(2005). Statistics for the Behavioral and Social Sciences: A brief Course (3e). Prentice Hall


        

" say it .....

self pity gets you nowhere
one must have the adventurous daring to accept oneself as a bundle of possibilities and undertake the most interesting game in world-making the most of one's best

[Henry Emerson Fosdick]