[Rasch] Data simulation
heene at psy.lmu.de
Sun Oct 7 06:12:26 EST 2007
Hello to everybody on the list,
I am currently planing a simulation study on some properties of the Rasch model. Unfortunatley, I can't find a straightforward way to simulate correlated error terms between dichotomous items in order to investigate different effects of model violations.
So finally I need to obtain matrices of dichotomous item responses with a priori specified degrees of violations of the local stochastic independence assumption/requirement.
Does anybody have an idea how this problem could be solved? Perhaps with Matlab, Mathematica, SAS, SPSS or something else?
Thanks in advance!
More information about the Rasch