[Rasch] Rasch person reliability and coefficient alpha

Stuart Luppescu s-luppescu at uchicago.edu
Sat May 31 06:14:23 EST 2008

Hi, I'm doing some survey analysis for an outside group. They gave me a
data set of 4 dichotomous items that go together; they say the items
have an alpha of .84. Yet when I Rasch analyze them, I get a reliability
of 0.00. The standard deviation of the person measures is 0.97 and the
RMSE is 1.14 (not surprising given we have only 4 pieces of information
for each person). That gives a separation of 0.00. 

The Rasch analysis results seem totally logical to me. Why is there such
a discrepancy between the two indices of reliability for the same data?
Stuart Luppescu -=- slu .at. ccsr.uchicago.edu        
University of Chicago -=- CCSR 
才文と智奈美の父 -=-    Kernel 2.6.24-gentoo-r8                
By failing to prepare, you are preparing to fail. 

More information about the Rasch mailing list