[Rasch] Simulated eigenvalues greater than empirical data analysis?
rmt at rasch.org
Sun Mar 11 15:11:41 EST 2012
Your first Table 13 indicates that your original Winsteps dataset is
probably unidimensional enough for practical purposes: First contrast,
eigenvalue = 1.9.
But Table 23 for the simulated data is definitely incorrect. Please email
me directly, and tell me the Winsteps version number (top of Winsteps output).
mike at winsteps.com
More information about the Rasch